Zenith Radio Corp. v. Matsushita Electric Industrial Co.

529 F. Supp. 866, 33 Fed. R. Serv. 2d 748, 1981 U.S. Dist. LEXIS 17920

Opinion

OPINION

EDWARD R. BECKER, District Judge.

Table of Contents

I. PRELIMINARY STATEMENT 872

A. Introduction 872

B. Case Management History as it Bears Upon Confidentiality 877

C. General Description of the Documents Whose Confidentiality is at

Issue 882

D. The Contentions of the Parties rj 884

E. The Framework for Dealing with the Motions and an Overview of

Their Resolution 887

II. THE PRINCIPLE OF CONFIDENTIALITY 889

A. Confidentiality Orders Pursuant to Federal Rule 26(c)(7) 889

B. The Validity of Pretrial Order 35 892

C. The Propriety of Wholesale Declassification ^93

III. COMMON LAW ACCESS RIGHTS TO JUDICIAL RECORDS 895

A. The General Nature of the Right 895

B. The Extent to Which Access Rights Attach to Judicial Records 897

1. Introduction 897

2. Discovery Materials 898

3. Material that is the Subject of an Evidentiary Ruling 898

4. Materials Referred to at a Hearing 899

5. Records Filed Under Seal that are the Ultimate Subject of a Dis-positive Ruling 901

C. Countervailing Interests: The Balancing Test 901

D. Access Rights to the Critical Documents at Issue 905

1. Introduction 905

2. The Document Submissions Sheets 905

3. The Final Pretrial Statement 906

4. Documents Referenced in the Summary Judgment Hearings and in the Final Pretrial Statement 907

5. The “Raw Economic Data” Contained in Sealed Discovery Responses and the Final Pretrial Statement Appendices 907

IV. FIRST AMENDMENT INTERESTS IN JUDICIAL RECORDS 908

A. Introduction 908

B. First Amendment Rights in Discovery Materials 908

C. Relationship Between the First Amendment and the Common Law Right to Inspect and Copy 913

D. First Amendment Rights to the Remaining Documents at Issue 914

V. CONCLUSION 915

I. Preliminary Statement

A. Introduction

This opinion addresses one of the most difficult and vexatious issues judges confront in managing complex litigation — the access and First Amendment rights of litigants and the public in connection with the large quantity of documents that is marked confidential pursuant to a protective order such as is entered to facilitate discovery in virtually every complex case. The setting of the present dispute is a massive international antitrust ease, the nature and the history of which are described in our opinions granting summary judgment for defendants on all of plaintiffs’ claims. See Zenith Radio Corp. v. Matsushita Elec. Indus. Co., 513 F.Supp. 1100 (E.D.Pa.1981), appeal pending, No. 81-2331 (3d Cir.) (Summary Judgment Motions Related to Plaintiff’s Sherman Act, Wilson Tariff Act, Clayton Act, and Robinson-Patman Act Claims) (hereinafter “Final Summary Judgment Opinion”); Zenith Radio Corp. v. Matsushita Elec. Indus. Co., 494 F.Supp. 1190, (E.D. Pa.1980), appeal pending, No. 80-2080 (3d Cir.) (1916 Antidumping Act). 1

Two cognate motions are before us, both rooted in pretrial order (“PTO”) 35 (Dec. 22, 1975), the “umbrella” confidentiality order governing pretrial discovery. 2 One motion seeks vacatur of PTO 35 in its entirety. The other, to which the bulk of this opinion will be devoted, seeks “wholesale” declassification, i.e., removal from confidential status, of: (1) all pleadings and other papers filed of record that have been stamped “confidential” under the aegis of PTO 35; (2) all documents offered by the parties at the pretrial evidentiary hearings; and (3) all documents produced in discovery which, under the aegis of PTO 35, have been stamped “confidential.” Defendants vigorously oppose the motions.

There have been to date over 3,500 docket filings in this case. These filings may be grouped into three categories relevant to the instant issues. The first category encompasses 385 filings that have been made under seal, pursuant to PTO 35. 3 The most

important of these filings is plaintiffs’ final pretrial statement [“FPS”], a multi-volume document that incorporates virtually all of the significant material contained in the other filings. Plaintiffs’ FPS contains some 17,000 pages and is cross-referenced to approximately 250,000 documents. The etiology of the FPS is explained in Part III of the Final Summary Judgment Opinion, 513 F.Supp. at 1130-1135. Suffice it to say here that the FPS provided the basic charter for our final summary judgment hearing. It is a document in which the plaintiffs 4 were required to set forth:

Each fact that the party intends to prove at trial either affirmatively or by way of defense, together with a list of: (1) the witnesses (including expert witnesses) whose testimony will be advanced to prove that fact; (2) the documents . . . which will be offered to prove that fact; and (3) line by line references to any portions of depositions and to answers to interrogatories and requests to admit which will be offered to prove that fact. 5

The second category sought to be declassified is comprised of many thousands of pages of documents, each assigned a document submission sheet [“DSS”] number, which plaintiffs offered at the pretrial evidentiary hearing. The evidentiary hearing was of enormous significance because plaintiffs’ case was founded almost entirely on documents. See id. at 1200-1202 (describing plaintiff’s litigation strategy). The pretrial evidentiary hearing, which lasted for over five weeks, had a two-fold purpose. First, it enabled us to rule upon the admissibility of the most critical documents in plaintiffs’ case (the DSS’s) so that we could determine whether we should consider them in ruling upon defendants’ summary judgment motions addressed to plaintiffs’ Sherman Act, Wilson Tariff Act, Clayton Act, and Robinson-Patman Act claims. Second, the hearing provided the forum for a pretrial ruling on the question whether the plaintiffs had come forward with a fair preponderance of evidence aliunde of the existence of a conspiracy among the various defendants to permit the introduction of co-conspirator declarations.

The third and final category of documents sought to be declassified consists of the millions of pages of documents produced in discovery. 6 This category includes the documents contained in the document depository created pursuant to PTO 219 to house what the parties (mainly the plaintiffs) asserted to be the more important documents in the case. No one has counted the number of documents in the depository, but we estimate it to be over 100,000. In light of the quantity of documents produced in discovery, the potential sweep of the wholesale declassification motion is enormous. Nor is much discussion required to demonstrate that the burden of reviewing each document individually in order to determine whether its continuing confidential status is justified would be almost impossible to meet.

PTO 35 was entered by our immediate predecessor in this case, Judge A. Leon Higginbotham, Jr., on December 19, 1975, slightly more than one year after the filing of the Zenith complaint. That order supplanted a more restrictive confidentiality order that had been entered by Judge Shaw in the District of New Jersey at the outset of the National Union Electronic Corporation (“NUE”) litigation. 7 PTO 35 states that many documents produced for discovery relate to trade secrets or other confidential research, development or commer cial information, as those terms have been defined pursuant to Rule 26(c)(7) of the Federal Rules of Civil Procedure. 8 The order defines “confidential information” as any information that is designated “confidential” at the time of its disclosure by the disclosing party. The order provides that confidential information shall be used only in the preparation for trial and/or trial of these actions, and not for any other purpose whatsoever.

PTO 35 further provides that confidential information may be disclosed only to persons denominated as “qualified persons,” and provides specific procedures by which additional “qualified persons” may be named. 9 The order requires parties objecting to a “confidential” designation of information to follow prescribed procedures in order to challenge that designation. Parties seeking confidential treatment of information, however, must ultimately bear the burden of establishing their right to the court’s protection. Upon final determination of these actions, PTO 35 requires all confidential information to be returned to the producing party. Of critical importance to the matters before us is the fact that PTO 35 governs only pretrial discovery and does not.address the confidentiality of documents at open pre-trial hearings or at trial.

The parties dispute whether PTO 35 was a product of their overt agreement, or at least their mutual acquiescence, 10 or whether Judge Higginbotham imposed the order after considering the parties’ proposals, but without their joint approval. It is indisputable, however, that: (1) throughout almost a decade of litigation, parties (plaintiffs as well as defendants) and non-parties to these actions have produced and marked “confidential” millions of documents in reliance upon PTO 35; (2) plaintiffs failed to challenge PTO 35 prior to June 28, 1977, by which time hundreds of thousands of documents and hundreds of voluminous pleadings had been designated confidential without objection; and (3) throughout the history of the litigation plaintiffs have never once availed themselves of the procedures of PTO 35 to challenge the propriety of the designation of any specific documents as confidential. Moreover, by designating as confidential hundreds of thousands of their own documents produced in discovery for defendants, plaintiffs have made as much use of PTO 35 as have defendants. 11 Under these circumstances, plaintiffs have ratified PTO 35 and we will treat it as a consent order.

When the wholesale declassification motion was filed on June 28, 1977, 12 this case was still assigned to the docket of Judge Higginbotham, who briefly considered the motion but deferred ruling on it in light of other, more pressing imperatives of case management. 13 Facing the onerous tasks of completing discovery, disposing of the surnmary judgment motions, and moving the case to trial, we followed essentially the same course. 14

We did, however, discuss declassification with counsel on a number of occasions shortly after assuming the case. See infra pages 877-878. More importantly in order to satisfy plaintiffs’ complaints that PTO 35 was hampering their trial preparation by barring them from adequately discussing material produced in discovery with Zenith executives and their consumer electronic products experts, we initiated procedures that expanded the scope of PTO 35 by creating two additional categories of persons entitled to have access to PTO 35 material: “qualified executives” and “qualified witnesses.” 15

Although we were (and still are) convinced that PTO 184 satisfied plaintiffs’ trial-preparation needs, plaintiffs continued to press for resolution of -their confidentiality motions, and we heard them again at the conclusion of the final summary judgment argument in August 1980. We again deferred consideration of these motions — this time pending our decision on the motions for summary judgment.

It is useful to underscore the different procedural settings of the case at the time of our various hearings on the declassification motions and at the present. The early hearings took place at a purely pretrial stage. The final hearing took place after the in limine pretrial evidentiary hearings that preceded the final summary judgment argument and which plaintiffs assert to have been part of the trial itself. We file this opinion in the wake of our grant of summary judgment in defendants’ favor as to all of plaintiffs’ claims, so that, unless that decision is reversed by the Court of Appeals, plaintiffs’ claims will not go to trial. 16

The grant of summary judgment as to plaintiffs’ claims does not moot either motion before us. In terms of the vacatur motion, we note that PTO 35 retains its vitality during the pendency of plaintiffs’ appeal, which could result in reversal of the judgment and consequent trial of plaintiffs’ claims. Moreover, PTO 35 governs the prosecution and defense of the defendants’ counterclaims which are still pending before us. Nor does the grant of summary judgment moot plaintiffs’ motion for wholesale declassification in view of plaintiffs’ persistent contention that: (1) the filings and documents in the case are or should be in the public domain, and the public should have access to them; and (2) the plaintiffs have the right to speak about them publicly.

The parties have presented us with multifaceted contentions regarding wholesale document declassification. We will presently set forth these contentions at length. We will also explain why the current motions will not be resolved wholly within the framework of the parties’ contentions, but rather within a some what, different framework which we will elaborate. Neither the contentions of the parties, however, nor the framework which we have designed for the disposition of these motions can be understood without some further background concerning the history of our case management as it bears upon the confidentiality of documents, and a fuller description of the nature of the critical documents at issue. We proceed now to supply that additional background.

B. Case Management History as it Bears Upon Confidentiality

Problems with document confidentiality first came to the court’s attention with the filing of plaintiffs’ motion for wholesale declassification. That motion was filed near the end of Judge Higginbotham’s stewardship. 17 Although Judge Higginbotham was unable to deal at length with the motions, he did make some telling observations and helpful suggestions in a bench memorandum during a pretrial conference on July 5, 1977:

A review of the briefs convinced me that among the massive documents produced and labeled as confidential in this case, there must be a plethora of such documents which no longer could be justified with the imprimatur of confidentiality as that term is related to discovery under the applicable cases. For such documents the reasons for confidentiality have expired with the passing of the seasons, and while the parties are obligated to treat these documents with total confidentiality until there has been some modification under Pretrial Order Number 35, the reasons for confidentiality have expired.

Yet, in contrast, I am equally confident that there are also many documents which were originally labelled confidential which are still confidential or at least, should not be automatically declassified by the rubric of ascertaining whether they were filed before or after September 21, 1974. As to the number which merit today confidentiality, I can only conjecture.

I am not unmindful that the Pretrial Order provides that any application to the Court for a ruling on confidentiality that the party seeking confidential treatment has the burden of establishing the right to such treatment. Yet, the problem at hand cannot be solved merely by drawing fast conclusions as to whether some burden of proof has been met in general for the burden of proof even in this argument has been met on some documents but not on others.

I received broad conclusionary statements about the confidentiality of documents which may vary in number from several thousands to hundreds of thousands. For this magnitude of discovery, conclusionary statements are not adequate to provide me with the specificity of contents of even one document or even as to one category of documents or as to any 'factual analysis of how the strands of categories are interrelated or intertwined in this discovery problem.

Though not the most rational method, perhaps the easiest judicial approach would be for the trial judge to peremptorily order the parties to file memorandum on each and every document which has been marked as confidential so that he can then make an independent determination as to whether there is still a viable claim of confidentiality, but if common sense is still a virtue, and if controlling unnecessary litigation costs is still an appropriate goal even in cases involving massive amounts of documents, there must be a more rational approach.

In this case it would be far more appropriate for each side to take an anatomical view of the case and to first identify those broad categories which are significantly different. Thus, which categories should have a period of confidentiality that precedes September 1974. They should note the time period for confidentiality of that category. As an example, should Category A because of its uniqueness have a period of confidentiality to 1973, or 1971, or 1969.

To the extent that the problems cannot be solved in terms of categories, then the Court will have no alternative but to deal with each document. The magnitude of such a challenge either through the Court or through a Master would be a clear revelation that all of us as lawyers and as a Court are functionally unable to work out rational alternatives to problems of moderate complexity. Therefore, I expect for counsel to give me a written report no later than ten a. m. July 18, 1977, as to whether this matter has been solved, and if in the unlikely probability that it has not been amicably solved, they must file their separate proposals as to each major category, identifying it with precision and the related time period for confidentiality of that category.

PTO 81, at 2-5 (July 5, 1977) (Transcript of Pretrial Conference).

Regrettably, Judge Higginbotham’s optimism was unrequited. As we have observed in earlier opinions, this case was then fraught with acrimony among counsel, and counsels’ efforts to agree on the kinds of categories suggested by Judge Higginbotham failed. When the declassification motion came to our attention we explored the possibility of designating categories of documents which might be declassified, but, we too were unsuccessful in attaining any agreement. Detailed consideration of the motions was less pressing than other exigencies of case management, so we deferred it.

As the case reached a more structured stage, with the filing of plaintiffs’ preliminary pretrial memorandum [“PPTM”] and defendants’ responsive PPTM’s, the spate of “confidential” designations continued. Indeed, most answers to interrogatories continued to be filed under seal, see supra note 2, and the PPTM, an assortment of briefs, some transcripts of pretrial conferences at which documents were discussed, and ultimately the FPS were likewise filed with confidential designations. Indeed the parties had long since gone far beyond erring on the side of caution and had stamped “confidential” on their submissions and discovery materials almost as a matter of course. 18 The inertia of habit ultimately carried the parties to such an extreme that even innocuous legal briefs began to sport the badge of confidentiality.

The keystone of our approach to case management was the FPS, which we have described above. The FPS’s detail was required principally as a surrogate for the plaintiffs’ inadequate responses to the defendants’ discovery requests. However, as we have explained in Part III of our Final Summary Judgment Opinion, 518 F.Supp. at 1130, the FPS fell short of its intended purpose. We were confronted at the time of its filing with defendants’ Motions for Summary Judgment which we had deferred until we possessed an adequate record upon which to base a decision. Rule 56(e) requires that a party opposing a motion for summary judgment “by affidavits or as otherwise provided in this rule . . . set forth specific facts showing that there is a genuine issue for trial.” The rule also requires summary judgment proceedings to be based upon admissible evidence. See id. at 1118 n.3. Mindful of those principles, we concluded early in 1980 that in order for us meaningfully to deal with the summary judgment motions, two devices would be required: (1) a pretrial evidentiary hearing, and (2) a summary judgment argument of sufficient length that plaintiffs would have an opportunity to call to our attention any document in the record that would create a genuine issue of material fact. We consequently proceeded to implement both devices.

The tenor of the pretrial evidentiary or “in limine” hearing has been described in opinion addressing the admissibility of public records and reports. See 505 F.Supp. at 1136-37. Suffice it to say here that plaintiffs marshalled their key documents, assigning each a DSS number, and argued the admissibility of most of them to the Court. Defendants argued in response, and extensive briefs were submitted by the parties. The hearing resulted in the three lengthy and detailed opinions mentioned above.

The in limine hearing commenced in June 1980. Not until the middle of the second week of the hearing was a question raised as to whether the hearing was open to the public or closed pursuant to PTO 35. Following a colloquy with counsel, see PTO 270, at 17-27 (Aug. 1, 1980), we ruled that the hearing would be considered open until consideration of the Yajima diaries, at which point the hearing would be closed, subject to a later review of the transcript. No one other than “qualified persons” had been in the courtroom up to that time. Presumably because of the ordinary character of the hearing, no spectators ever appeared for the balance thereof. Hence, no one was ever excluded from the courtroom. 19

There is another, and equally important, facet of the matter: when the transcripts of the hearing were filed in July 1980, they were not marked “confidential” and they were not sealed. They have remained open and on the public record, available for inspection by anyone who wishes to see them. Moreover, counsel have been aware that the transcripts were filed unsealed, for the cop ies of the transcript volumes sent to them almost contemporaneously (daily copy was furnished) were not marked “confidential,” as was the custom when PTO 35 had been invoked. No one, however, has ever moved that the transcripts be sealed.

Our opinions addressing the evidentiary questions arising during the hearings contain detailed discussions of the documents at issue. The opinions were, of course, immediately filed on the public record and were promptly published. It should also be noted that almost all of the fifty-nine briefs filed within the ten week period surrounding the pretrial evidentiary hearings, most of which were filed by defendants and many of which dealt with documents considered during the hearings, were not filed under seal. 20

Argument on the summary judgment motions lasted seven days. As is made clear by PTO 154, supra note 5, by the transcripts of numerous pretrial conferences addressed to case management, and the transcript of the summary judgment hearing the FPS was the principal focus of the summary judgment hearing. The critical question before us was whether there was anything in the FPS, the distillation of thirty million document pages and ten years of litigation, that created a genuine issue of material fact as to the existence of a Sherman Act conspiracy among the defendants. The confidential documents rescribed or referenced in the FPS were a constant subject of colloquy in the courtroom. For the most part, they were simply paraphrased, summarized, or referred to, though on occasion excerpts were read into the record. The public was never excluded from the hearing, and no request for exclusion was made. The briefs filed by defendants addressing the summary judgment issues were, for the most part, filed on the public record. The transcript of the summary judgment hearing was also filed on the public record. Counsel were aware of that fact from the time of filing of the transcript, shortly after the hearing, and have never voiced any objection. Our summary judgment opinion discusses the contents of the FPS in detail. That opinion is of course on the public record and has been published.

We have noted above that at no time did plaintiffs utilize the procedures of PTO 35 to object to the “confidential” classification of any specific documents, and that their only attack has been via their motion for wholesale declassification. Concomitantly, the defendants have never been put to the burden of justifying confidential treatment for any given document. Defendants’ response has been to object to wholesale declassification and to point out in general how declassification of documents containing technical, financial, and marketing information would be harmful to their interests, even where the data is as much as ten years old.

The vast bulk of defendants’ response has been by means of defense counsels’ oral and written submissions. In 1977, however, while the matter was pending before Judge Higginbotham, the Hitachi defendants filed a lengthy telex from a Hitachi Corporate Secretary, Mr. Miyamoto, which Judge Higginbotham accepted in lieu of an affidavit. The telex explained, in general terms, how declassification would be harmful to Hitachi’s interests. 21 A similar submission in affidavit form was made at that time by Mr. Kenji Yamagishi on behalf of Matsushita. 22

The judges who have presided over this ease have had a number of occasions to act upon confidentiality matters. Some of these occasions arose when the parties submitted stipulations that relief from PTO 35 be granted in response to a government agency demand. There were also several contested confidentiality motions. We annotate these instances briefly to amplify the background of the treatment of confidentiality matters in this litigation and to contrast them with the parties’ failure to challenge specific documents.

Judge Higginbotham’s major confidentiality problems related to whether the fruits of discovery, en masse, could be turned over to the United States International Trade Commission, 23 and whether documents produced to the ITC by a third party could be subpoenaed by parties in this action. 24 Shortly after the assignment of the case to our docket, we executed a consent order permitting the United States Department of Justice to examine selected fruits of discovery. 25 We also had occasion to deal with a motion to depose Bernard Nash of the Washington law firm of Blum, Nash and Parker, a lawyer or lobbyist for Zenith, and to obtain access to his law firm’s records on the theory that Zenith had breached PTO 35 by providing Nash with protected documents which Nash in turn allegedly furnished to the United States Senate Antitrust and Monopoly Subcommittee. 26

C. General Description of the Principal Documents Whose Confidentiality is at Issue

We turn now to a general description of the most critical documents and filings in the case, with a view ultimately to determining the need for their continued confidentiality. For reasons that will appear, it is only these documents and filings which we will consider. Because we have already described the pretrial evidentiary hearings, it will be convenient to begin with the DSS’s. The important documents considered during the evidentiary hearings fall into a number of broad categories, as follows:

1. Documents, including certain findings, promulgated by the United States Treasury Department and the United States Tariff Commission in connection with proceedings under the 1921 Antidumping Act.

2. Documents, including certain findings, promulgated by the United States Tariff Commission and its successor, the United States International Trade Commission, as well as the Secretary of Labor, under §§ 301(b)(1) and (c)(1) and (2) of the Trade Expansion Act of 1962 and §§ 201(b) and 221 of the Trade Act of 1974.

3. Certain purported findings and related documents of the Japanese Fair Trade Commission (“JFTC”) arising out of proceedings in two cases before the JFTC: the first, brought in 1957 against the Home Electric Appliance Market Stabilization Council, some of whose members are defendants in this action, alleging industry-wide price-fixing; and the second, brought in 1967, alleging retail price maintenance against defendant Matsushita Electric Industries Company.

4. Diaries of officials of several of the Japanese defendants, alleged to contain evidence of the conspiracy asserted in plaintiffs’ complaint, which were seized in 1966 and utilized in the course of JFTC Case No. 6 of 1966 (the “Six Company Case”). That case involved an investigation and proceedings against six of the defendants: Sanyo Electric Corporation, Tokyo Shibaura Electric Corporation (Toshiba), Hayakawa Electrical Industry Corporation (now Sharp Corporation), Hitachi Industry Corporation, Matsushita Electric Industry Corporation (MEI), and Mitsubishi Electric Corporation (MELCO). The six companies were charged with a conspiracy to fix prices and to engage in a variety of activities in violation of the Japanese Anti-Monopoly Law. The eight diaries in dispute have been attributed to Mr. Yajima (3 diaries), an employee of Toshiba; Messrs. Yamamoto (2 diaries) and Yamada, both employees of Hitachi; Mr. Okuma, an employee of MEL-CO; and Mr. Tokizane, an employee of MEI. Also included in this category are a number of internal company memoranda which were seized by the JFTC.

5. Transcripts of testimony and of protocols by witnesses in the Six Company case. A protocol is a written statement outlining the substance of oral discussions between a JFTC staff member and a witness, which the witness signs to indicate that the contents comport with his statement.

6. Various agreements and rules of certain Japanese manufacturers’ associations relating to export practices.

7. Various documents alleged to be minutes or memoranda of meetings of certain Japanese manufacturers’ associations, including the Electronics Industries Association of Japan (EIAJ) and the so-called TV Export Council.

8. A purported internal memorandum of the Japan Victor Company, fifty-one percent of which is owned by MEI, allegedly reflecting the decision made by the EIAJ to conceal from the Japanese Ministry of International Trade and Industry (MITI) the discrepancy between domestic and export prices and suggesting changes in accounting methods by which such concealment could be accomplished.

9. Numerous memoranda, letters, telexes, and transactional documents produced by the defendants in discovery, involving the Japanese manufacturers, their trading companies, their American sales subsidiaries, and various American customers, which, in plaintiffs’ submission, show a pattern of “under the table” or concealed rebates that reduced the price of Japanese television receivers to American customers below the “check price” that was reported to United States Customs, and also reveal a “cover-up” of what plaintiffs describe as a predatory export scheme.

10. A potpourri of other documents, produced during discovery for the most part from defendants’ files.

11. Voluminous reports setting forth the opinions and supporting data of plaintiffs’ experts. The reports at issue were: (1) “Economic Study of the Japanese Television Industry,” by Dr. Horace J. DePodwin, Dr. David Schwartzman, and Marcio Teixeira of Horace J. DePodwin Associates, Inc., an economic consulting firm; (2) “The Pervasive Use of Collusive and Company Group (Keiretsu) Activities in Achieving the Rapid Increase of Japanese Exports of Television Receivers to the United States,” by Professor Kozo Yamamura, Chairman of the Japan Studies Program and Professor of Economics and East Asian Studies at the University of Washington; (3) “Economic Analysis of Evidence Relating to Japanese Electronic Products Antitrust Litigation,” by Stanley Nehmer of Economic Consulting Services, Inc.; (4) “The Impact of Japanese Financial and Employment Practices on Japanese Production, Marketing, and Price Behavior,” by Professor Gary R. Saxon-house, Professor of Economics at the University of Michigan; and (5) “Vertical Restraint by Japanese Television Manufacturers: Anticompetitive Effects,” by Professor John Owen Haley, Associate Professor of Law at the University of Washington.

We turn from the DSS’s to those critical filings on the record of the case that are designated “confidential.” The most critical filing in plaintiffs’ case is, as we have stated, the FPS. The FPS incorporated the DSS’s, but also subsumed numerous other matters, almost all document-related. The subject matter of these other documents is covered in Part VII of the index to the Final Summary Judgment Opinion, 513 F.Supp. at 1115-16. While these other documents are more numerous, in qualitative terms they are similar to the DSS’s.

One facet of the FPS and of several of the confidential record filings bears special mention. Included in the appendices to the FPS are the voluminous model-by-model matchups and price comparisons described in the 1916 Antidumping Act Opinion, 494 F.Supp. at 1204. Also included among the sealed filings are the Answers to Interrogatories 45 and 46(c) of Plaintiffs’ Interrogatories to Defendants, Set No. 1, which set forth the prices of defendants’ television receivers at each level of distribution, and several other similar interrogatory responses. All of this economic data has been analyzed and distilled by plaintiffs’ experts in their reports. It is plain, however, that if there is any potential competitive harm to defendants from declassification, it would come from disclosure of this raw economic data, not from the expert reports or the FPS text. We do not take up the other record filings, voluminous as they are, because they are essentially subsumed within the FPS. 27

We have thus generally described the critical documents. What is important of course is the question whether the documents — and the FPS is principally an agglomeration of documents and excerpts therefrom — are of a character that their confidentiality must be protected under the applicable law. The resolution of that question requires more than a general description of the documents; it requires an analysis of their underlying content, in the wake of a description of the applicable law.

D. The Contentions of the Parties

Plaintiffs’ motions are grounded on a number of arguments. First, they maintain that most of the documents designated by defendants as confidential do not qualify as trade secret or other confidential research, development, or commercial information within the meaning of Federal Rule 26(c)(7) which is, as we have noted, the provision that informs PTO 35 for purposes of this case. Even if Rule 26(c)(7) does cover the information, say plaintiffs, there is not “good cause” to protect them from disclosure. In plaintiffs’ submission, the material at issue, by and large, never qualified under Rule 26(c)(7), and if it ever did, no longer qualifies because the passage of time has rendered it stale. Much of the technical and commercial data at issue is from ten to twenty years in age.

Second, although the pretrial evidentiary hearing was closed in part to the public, plaintiffs submit that the closure was improper and, therefore, that all the documents offered at the hearing, regardless of our admissibility rulings, must be declassified because of the common law right of access in the public to inspect and copy court documents. This result is also compelled, in plaintiffs’ submission, by the reference to many of the documents in our various published opinions in this case. Plaintiffs point out that this case has received widespread publicity in domestic and international legal and financial circles, and that it is a case of great public importance. Plaintiffs contend that legal scholars and economists cannot evaluate the result we have reached unless they have access to the “raw material” upon which our decisions are based. This includes, plaintiffs say, at least the following: (1) all the DSS’s offered and considered by us in the pretrial evidentiary hearing; 28 (2) all the documents referred to during the summary judgment hearing; (3) all documents referred to in the Final Summary Judgment Opinion; and (4) plaintiffs’ entire FPS and the 250,000 documents referenced therein. 29 Plaintiffs seek to justify inclusion of this latter category by emphasizing that the FPS was the focal point of the summary judgment hearing and deliberation, and of the opinion granting summary judgment for defendants.

Finally, plaintiffs claim that, PTO 35 notwithstanding, the First Amendment gives them the right to. speak about and, upon request, to disseminate material obtained in discovery. Plaintiffs’ position on this last point relies, for the most part, upon In Re Halkin, 598 F.2d 176 (D.C.Cir.1979), a controversial decision addressing the question of First Amendment rights in discovery materials.

Defendants counter plaintiffs’ arguments on numerous grounds. First, they argue that PTO 35 was a typical confidentiality order, eminently proper in a case such as this where, pursuant to plaintiffs’ massive discovery requests, the defendants and a number of third parties produced literally millions of pages of documents containing detailed technical and commercial information exposing the very heart of their business operations. Defendants submit that PTO 35 has effectuated the orderly process of discovery and pretrial preparation in this litigation. They point out that for a decade both parties and non-parties, in good faith and in reliance on PTO 35, have produced countless documents the discovery of which they might have challenged on a document-by-document basis but for PTO 35. They argue that those who have made disclosure would be seriously and irreparably harmed if confidentiality were now revoked.

In defendants’ submission, a competitor would have a distinct advantage in planning the manufacturing and marketing of its own products if it were armed with information about defendants’ production and sales plans and techniques; marketing, pricing, and advertising plans and philosophy; customer information, long term plans, product development patterns; and other vital competitive information. Defendants claim that the materials previously disclosed by them in this case contain such information. 30 Moreover, defendants claim that wholesale declassification would be grossly unfair in view of plaintiffs’ failure timely to avail themselves of the procedures afforded by PTO 35 for challenging the confidential designations.

As a major predicate for these contentions, defendants point to their production to the plaintiffs of copies of their pricing materials, which demonstrate the prices of their merchandise at various levels of distribution. They then explain that the pricing process does not vary greatly from year to year, and the conclusion which they thereupon draw is ■ that much can be learned about the manner in which the defendants arrive at their prices by analyzing this data, even if it is relatively “stale,” and by comparing any year’s price list at various sales levels. Defendants note too that marketing strategies, and the amount of money allocated for promotional needs, necessary discounts, and allowances, remain relatively constant from year to year. They allege that study of these materials, regardless of the year chosen, would enable a skilled competitor to understand the pricing and marketing patterns of defendants and plan its own operations accordingly.

The Japanese manufacturing defendants have produced to the plaintiffs various materials documenting their transactions with certain private label purchasers, such as Sears Roebuck and Company and J. C. Penney. These documents demonstrate the volume of purchases by these private label retailers, the types of models of consumer electronics products which these customers purchase from defendants, and the price of these products. Although Zenith has not heretofore sold to private label retailers, there is some indication in the record that Zenith may be interested in this type of business in the future. In view of this fact, defendants evince concern that the private label materials may be used to Zenith’s commercial advantage, even to the extent of shedding light upon the negotiating processes by which defendants arrive at these agreements. Of equal if not greater concern to defendants is the possible consequence of exposure of such data to the private label retailers themselves, in that customer relationships could be impaired if these retailers knew the terms on which a manufacturer sold to the retailer’s competitors.

A number of defendants have produced patent and licensing agreements between individual defendants and other third parties. These agreements provide such information as the cost of the licensed technology and data regarding production quantities and the patented technology itself. Zenith, as well as others, competes for the sale of such licensed technology and, in defendants’ submission, such information might conceivably be helpful to Zenith’s personnel in marketing Zenith’s competing technology. Furthermore, claim defendants, some of these agreements provide that defendants will not disclose this information to any third parties. Defendants assert that if this information were to be placed in the public domain, or if Zenith’s key executives were to be allowed unfettered access to these documents, defendants would suffer the disruption of advantageous business relationships as well as the possibility of suit for breach of contract.

Taking still another approach to the problem, defendants allege that PTO 35 was and is needed because Zenith has on occasion used discovery material to fuel a public relations campaign and lobbying efforts in the Congress and executive agencies against the Japanese consumer electronics products manufacturers and exporters. Defendants claim that Zenith has made use of such information in the preparation of submissions to government agencies such as the Antitrust Division of the Justice Department, the International Trade Commission, the Senate Antitrust and Monopoly Subcommittee, and the press. In defendants’ words:

These submissions, carefully phrased with self-serving statements that they do not disclose confidential information, contain artfully edited portions of information contained in materials obtained in this litigation from defendants, the end effect of which is to present to these various bodies a constant flow of half-truths and misinformation in a deliberate anticompetitive campaign to support Zenith’s allegations that Japanese competition is destroying its business. 31

Finally, defendants assert that the in limine hearing was properly closed, because an open hearing would have exposed sensitive commercial information, properly protected by Rule 26(c)(7), from the scrutiny of those who might use it to defendants’ competitive disadvantage. They also dispute plaintiffs’ legal contentions about the common law right of access to documents and First Amendment rights in discovery materials. Thus, defendants vigorously object to vacatur of PTO 35 and wholesale declassification. 32

E. The Framework for Dealing with the Motions and an Overview of their Resolution

Having set forth in some detail the parties’ approaches to the problems raised by the pending motions, we now turn to an elaboration of the framework within which we will decide them. As will be seen, we conclude that we need not consider the vast bulk of the documents with which the parties’ submissions are concerned. Rather, we will confine our attention to the most important categories of documents and filings. In this segment we will outline our approach to the legal issues and give an overview of our ultimate resolution of the factual questions. In the following sections, we will discuss the principles of law that fill in the contours of this decisional framework, and apply those principles to the facts at bar.

In view of the enormous number of documents put in issue by these motions, our first order of business must be to determine whether wholesale declassification is appropriate. That deliberation presupposes a judgment as to the validity of PTO 35, however. We will therefore address the appropriate standard for confidentiality orders issued pursuant to Rule 26(c)(7). 33 We will conclude that PTO 35 is a valid order, one well within the scope of our discretion under the Federal Rules.

Having concluded that PTO 35 is valid, we will decline to unseal en masse the documents at issue. The motion for wholesale declassification is made under circumstances in which the parties have agreed to the entry of a blanket protective order. The court cannot consider individually more than a small number of the documents at issue, though many of these documents contain valuable and sensitive commercial information which, if disclosed, would damage the economic interests of their owners. We will conclude that an application for wholesale declassification of documents is beyond the scope of the appeals process contemplated by PTO 35, and that the parties waived their right to make such an application by entering the consent order. In any event, public and private interests in the general and efficient management of complex cases militate against granting plaintiffs’ motion. 34

Although we deny plaintiffs’ motion for wholesale declassification as such, we will treat their motion as one seeking declassification of the critical categories of documents in the case. Thus, we must turn to the question of what filings are on the judicial record, or, more precisely, to identification of the judicial records to which the common law access rights of the public attach. In this regard, we will first discuss the general nature of the common law right to inspect and copy judicial records. We will then discuss rights of access to discovery materials, material that is the subject of an evidentiary ruling, materials referred to at a hearing, and records filed under seal that are the ultimate subject of a dispositive ruling (such as our final summary judgment ruling in this case). We will then review the nature of interests that may militate against disclosure. Finally, we will apply these principles to the critical documents at issue in this case.

Turning first to the DSS’s produced at the in limine hearing, we will explain why we deem all the documents considered during the hearing, whether or not ruled admissible, to be declassified. 35 This conclusion results in the declassification of the JFTC record, already largely open in Japan, plaintiffs’ expert reports, and a host of other materials. 36 Defendants and third parties may suffer some harm as a result of the exposure of these materials, but there has been no showing of potential harm sufficient to overcome the presumption of public access.

Turning to the filings on the court record, we are satisfied that the interests of the parties can be accommodated by limiting our consideration to two categories of documents: (1) the raw economic data contained in such documents as defendants’ answers to interrogatories 45 and 46c of plaintiffs’ first set of interrogatories and their model-by-model price comparisons; and (2) the FPS. The FPS subsumes and incorporates in detail all of plaintiffs’ findings of any significance other than the raw economic data. 37 We will conclude that common law access principles do not require release of the raw economic data filed of record. Our final and major concern thus will be the declassification of the FPS, its appendices, and the documents referred to therein.

We start with the propositions that the summary judgment argument was open to the public, and that the FPS was the basic charter for that hearing. More than seven days of oral argument, dozens of briefs running into the thousands of pages, and our lengthy Final Summary Judgment Opinion were devoted essentially to a determination whether the FPS contained evidence creating a genuine issue of material fact that would defeat summary judgment. For reasons that will appear, we will declassify the FPS, and those of its appendices that do not contain economic data. We will not, however, declassify the documents referenced in the FPS and the other appendices. Although there is matter contained within the FPS that might legitimately remain classified as confidential, on balance we think that protection is not warranted. As will be seen, part of our rationale for declassifying the FPS is our sense of the duty of the courts to explain their judgments and to render them understandable by those who might wish to evaluate the outcome of a given case.

Because we conclude that common law principles do not require access to the raw economic data, including some of the FPS appendices, and the documents referred to in the FPS and at the summary judgment hearing, we will also consider whether First Amendment principles require disclosure. 38 We will take up first the difficult question of what First Amendment rights inhere in discovery materials and next the relationship between the First Amendment and the common law right to inspect and copy. Though we conclude that discovery materials are entitled to some constitutional protection, applying the legal principles to the facts of the case at bar, we conclude that First Amendment interests do not justify further declassification.

In sum, we will deny plaintiffs’ motions for vacatur of PTO 35 and for wholesale declassification. We will declassify the DSS’s, the body of the FPS, and the non-economic appendices thereto. We will decline to unseal the remaining categories of documents, including the raw fruits of discovery.

We will now commence our discussion of the applicable law. The several pertinent lines of cases are very much interrelated, almost as though the several strands are wound together to form a skein.

II. The Principle of Confidentiality

A. Confidentiality Orders Pursuant to Federal Rule 26(c)(7)

The propriety and desirability of protective orders securing the confidentiality of documents containing sensitive commercial information that are the subject of discovery in complex cases is too well established to belabor here. We are unaware of any case in the past half-dozen years of even a modicum of complexity where an umbrella protective order similar to PTO 35 has not been agreed to by the parties and approved by the court. Protective orders have been used so frequently that a degree of standardization is appearing. The Manual for Complex Litigation contains a representative umbrella protective order form. Board of Editors, Federal Judicial Center, Manual for Complex Litigation, app. § 2.50, reprinted in 1 Moore’s Federal Practice (pt. 2) app. § 2.50, at 316-19 (1981). And the course materials in virtually every continuing legal education seminar on civil litigation contain a variety of umbrella protective order forms. These forms are invariably grounded upon Rule 26(c)(7) of the Federal Rules of Civil Procedure. 39

We have already explained why we think that PTO 35 is a consent decree entered pursuant to the court’s authority under Rule 26(c). In this section we will examine the validity of PTO 35 under Rule 26(c), treating PTO 35, for analytic purposes, as an order imposed by the court on the parties. If PTO 35 is valid within this framework, then it is also valid when viewed as a consent agreement approved and given force by the court. 40

Analysis of protective orders under Rule 26(c)(7) requires three lines of inquiry. First, is the matter sought to be protected “a trade secret or other confidential research, development, or commercial information” which should be protected? Second, would disclosure of such information cause a cognizable harm sufficient to warrant a protective order? Third, has the party seeking protection shown “good cause” for invoking the court’s protection? In this case, defendants allege that the materials at issue should not be disclosed because the commercial information contained in them would put them at a serious competitive disadvantage. In that context we are called upon to decide whether such commercial information is properly protected, whether the alleged competitive disadvantage from disclosure is a kind of injury cognizable under Rule 26, and whether defendants have shown good cause. We discuss the law applicable to these inquiries seriatim.

The least troubling inquiries focus upon the first two points: confidential commercial information and competitive disadvantage. The cases concerning the protection of trade secrets and confidential commercial information, decided under Rule 26(c)(7) and its predecessor, Rule 30(b) 41 indicate that this subject-matter category is broad enough to include a wide variety of business information, including the kinds of matters sought to be protected by defendants. 42

Competitive disadvantage is a type of harm cognizable under Rule 26. Although the cases have not tried to identify competitive disadvantage as any one type of the three specific harms — annoyance, embarrassment or oppression — mentioned in Rule 26 which might subsume this harm, it is clear that a court may issue a protective order restricting disclosure of discovery materials to protect a party from being put at a competitive disadvantage. See, e.g., Parsons v. General Motors Corp., 85 F.R.D. 724, 720 (N.D.Ga.1980); Vollert v. Summa Corp., 389 F.Supp. 1348 (D.Hawaii 1975); Maritime Cinema Serv. Corp. v. Movies En Route, Inc., 60 F.R.D. 587 (S.D.N.Y.1973); Borden Co. v. Sylk, 289 F.Supp. 847 (E.D. Pa.1968), appeal dismissed, 410 F.2d 843 (3d Cir. 1969).

The party seeking a protective order bears the burden of showing good cause for the order to issue. Reliance Ins. Co. v. Barron’s, 428 F.Supp. 200 (S.D.N.Y.1977); Davis v. Romney, 55 F.R.D. 337 (E.D.Pa. 1972); Hunter v. International Sys. & Controls Corp., 51 F.R.D. 251 (W.D.Mo.1970); Essex Wire Co. v. Eastern Elec. Sales Co., 48 F.R.D. 308 (E.D.Pa.1969). In order to establish good cause, it must be shown that disclosure will work a clearly defined and serious injury, Essex Wire Co., supra; United States v. Lever Bros. Co., 193 F.Supp. 254 (S.D.N.Y.1961), cert. denied, 371 U.S. 932, 83 S.Ct. 310, 9 L.Ed.2d 272 (1962), 43 and that the party resisting disclosure “will indeed be harmed by disclosure.” Johnson Foils, Inc. v. Huyck Corp., 61 F.R.D. 405, 409 (N.D.N.Y.1973). Accord, Reliance Ins. Co., supra.

It has been held that in order to show good cause, the injury which allegedly will result from disclosure must be shown with specificity, and that conclusory statements to this effect are insufficient. United States v. Hooker Chem. & Plastics Corp., 90 F.R.D. 421 (W.D.N.Y.1981); Rosenblatt v. Northwest Airlines, Inc., 54 F.R.D. 21 (S.D. N.Y.1971); Hunter, supra; Technical Tape Corp. v. Minnesota Mining & Mfg. Co., 18 F.R.D. 318 (S.D.N.Y.1955). It has also been held that the specific instances where disclosure will inflict a competitive disadvantage should be set forth in more than the briefs or the hearsay allegations of counsel’s affidavit, for a protective order should not issue on that basis alone. See Reliance Ins. Co. v. Barron’s, supra; Rosenblatt v. Northwest Airlines, Inc., supra; Apco Oil Corp. v. Certified Transp., Inc., 46 F.R.D. 428, 432 (D.Mo.1969); Paul v. Sinnott, 217 F.Supp. 84 (W.D.Pa.1963). We think, however, that hard and fast rules in this area are inappropriate. Frequently the injury that would flow from disclosure is patent, either from consideration of the documents alone or against the court’s understanding of the background facts. The court’s common sense is a helpful guide.

An attempt to show that disclosure will indeed work a competitive disadvantage might be undermined if the information sought to be protected were stale. There is a paucity of case law in the area. In Vollert v. Summa Corp., supra, and Hecht v. Pro-Football, Inc., 46 F.R.D. 605 (D.D.C. 1969), information up to three years old was held entitled to confidentiality and a court’s attendant protection. On the other hand, in United States v. International Business Mach. Corp., 67 F.R.D. 40 (S.D.N.Y.1975), information three to fifteen years old was held not entitled to protection because, in the court’s view, it revealed little, if anything, about the contemporary operations of the party resisting disclosure. For the same reason, United States v. Lever Bros. Co., supra, held that information three to eight years old should not be protected. Indeed, in Rosenblatt v. Northwest Airlines, Inc., supra, the need for a court’s protection was held to be diminished because the information sought to be protected was one year old.

Notwithstanding the conclusions in these cases, it is terribly difficult to establish, on any principled basis, temporal boundaries governing the protection to be accorded information. While at first blush one might doubt that harm could be caused by the disclosure of stale information, there is sense in the argument, which defendants urge, that old business data may be extrapolated and interpreted to reveal a business’ current strategy, strengths, and weaknesses. 44 It would appear that, in the hands of an able and shrewd competitor, old data could indeed be used for competitive purposes.

We also conclude that PTO 35 comes well within the bounds of discretion established by Rule 26. By its terms, PTO 35 applies only to material protectible under Rule 26— materials which “relate to trade secrets, or other confidential research, development or commercial information, as such terms have been defined pursuant to Rule 26(c)(7) of the Federal Rules of Civil Procedure.” PTO 35 does not shift the burden of proof, but requires that, upon objection, the party electing to classify information justify its action pursuant to the Federal Rules of Civil Procedure. Finally, PTO 35’s appeals process insulates the order against the danger that its provisions may sweep too broadly. Any party objecting to a confidentiality classification may request a determination of a particular document’s entitlement to protection under Rule 26(c).

Finally, a protective order should always be narrowly drawn. Two considerations mandate this constraint. First, an overbroad protective order may constitute an abuse of discretion by exceeding the authority granted by Rule 26. Cf. Gulf Oil Co. v. Bernard, 452 U.S. 89, 101 S.Ct. 2193, 68 L.Ed.2d 693 (1981) (order that banned all communications, without prior approval of court, concerning class action between parties or their counsel and any actual or potential class member who was not a formal party, entered without findings of fact or explanatory opinion, exceeded bounds of discretion under Rule 23(d)). Even when a court acts within the bounds of discretion set by Rule 26(c), its actions should be informed by the second consideration mandating a narrowly drawn order, the First Amendment overbreadth doctrine. This doctrine, along with the other First Amendment issues pressed upon us by plaintiffs, is considered in detail in Part IV, infra.

B. The Validity of Pretrial Order 35

We find that PTO 35 is valid under Rule 26(c)(7). We reach that conclusion because: (1) the material that it protects is confidential commercial information; (2) the harm that it seeks to prevent is cognizable under Rule 26(c); and, (3) both at the time it was entered and at the present time, defendants (as well as plaintiffs) have shown good cause for the protective order to issue. It is plain from a reading of both the NUE and Zenith complaints, which spanned the law of antitrust and focused on defendants’ price behavior, that large quantities of sensitive commercial data would be sought in discovery. That prospect in fact materialized, as our description of some of the price data generated in discovery makes clear. Thus, the propriety of some form of umbrella protective order was never seriously in doubt.

The affidavits filed by defendants to support continued enforcement of PTO 35 45 do not detail the harm which would result from disclosure on a document-by-document basis. Rather, they describe the harm which would result from the disclosure of discrete categories of information. We find that the affidavits are consistent with our knowledge of the material at issue. We are also satisfied that by describing the harm which would result from the disclosure of categories of information defendants have shown good cause. . In reviewing other cases in which a protective order has issued, 46 we find that those courts discussed categories of information such as customer, supplier, and price lists; financial records; and contract terms; the very categories present here. Only rarely were individual documents subjected to a Rule 26(c) analysis. This case is distinguished from the usual Rule 26(c) case only in that the categories involved here contain much, much larger numbers of documents. Grouping huge amounts of cumbersome data into manageable categories for the purpose of supporting a Rule 26(c) order is desirable from the standpoint of case management and is consistent with the instruction of Rule 1 that the Rules of Civil Procedure shall be construed to secure the “just, speedy, and inexpensive determination of every action.”

The need for PTO 35 is hot diminished by our disposition of defendants’ summary judgment motion. We have defendants’ counterclaims still before us for eventual trial. Moreover, our summary judgment order may not have undercut PTO 35’s vitality as it pertains to plaintiffs’ case, for our decision may be reversed and plaintiffs’ case may yet go to trial.

Plaintiffs’ motion for vacatur will not be granted on the ground that PTO 35 exceeds our authority under Rule 26(c), for we have found the contrary to be true. Still unanswered is the question whether wholesale declassification per se is proper in this case. This issue raises issues distinct from, albeit related to the Rule 26(c) determination. To those issues we now turn.

C. The Propriety of Wholesale Declassification

In this section we examine the implications of and problems posed by plaintiffs’ motion for wholesale declassification. The phenomenon of declassifying as a group all of the documents protected by an umbrella protective order raises special problems that we believe must be addressed separately from the other issues raised in this case.

The underlying assumption of plaintiffs’ motion is that the usual procedures under PTO 35 for classifying and challenging the classification of documents may be translated intact into the context of wholesale declassification. In other words, plaintiffs contend that because they have challenged the need for confidentiality of all the protected documents, defendants must now meet their burden of showing good cause for the confidential designation of all the documents just as if plaintiffs had challenged the classification of a single document.

We believe, however, that the proposed procedure would effect an unintended metamorphosis of PTO 35, for there is a qualitative difference between sustaining the confidentiality of a single document at the time of its designation and sustaining the confidentiality of hundreds of thousands of documents en masse years after they have been produced and given a confidentiality designation.

Wholesale declassification cannot meaningfully be attained unless the necessary determinations can be made by some sweeping rule of thumb, e.g., documents over “x” years of age are conclusively ineligible for continued protection. But, as we have explained, rules of thumb are particularly inappropriate in this context. We conclude that wholesale declassification cannot rationally be obtained without reviewing each of the countless separate documents or at least the many discrete categories which constitute the mass of protected documents.

Wholesale declassification of millions of pages of documents, years after they have been produced, would require a tremendous concentration of judicial and litigants’ resources. It would be virtually impossible for us to consider declassification document-by-document at this juncture. PTO 35 provided a procedure for contemporaneous objection to confidentiality designations which was readily available to plaintiffs, but which they chose not to exploit. Although PTO 35 provides that the failure of any party to challenge a confidentiality designation contemporaneously shall not be deemed a waiver of its right to challenge the propriety of such designation at any time thereafter, this does not mean that a party may sit on its hands while the mountain of discovery materials grows and then attempt to challenge the protection of such material with the same ease with which it could have raised an objection contemporaneously. A party seeking wholesale declassification must first attempt to justify the investment of judicial and private resources demanded by such an exercise. Thus, to this extent, wholesale declassification shifts part of the burden to the party seeking disclosure. Cf. Iowa Beef Processors, Inc. v. Bagley, 601 F.2d 949 (8th Cir.), cert. denied sub nom. Iowa Beef Processors, Inc. v. Smith, 441 U.S. 907, 99 S.Ct. 1997, 60 L.Ed.2d 376 (1979) (before dissolving longstanding protective order, party seeking dissolution bears burden of showing that intervening circumstances have removed potential prejudice from disclosure that protective order was initially intended to prevent).

In terms of complex case management, we also believe that wholesale declassification is a poor, inappropriate, and unfair tool, not only with respect to the interests of the litigants, but also with respect to the interests of third parties who are frequently drawn into the vortex. It is common that, in response to a subpoena to attend a deposition and produce related documents, and in reliance upon an umbrella confidentiality order, third parties divulge sensitive commercial information without asserting their right to be protected by a court in their “home jurisdiction.” 47 Wholesale declassification might unfairly prejudice the rights of such third parties because they may not be in a position to contest the wholesale declassification motion in which their documents are implicated. This prejudice may result because they are removed from the controversy or difficult to locate years after their submission of documents or because they lack the resources or will to oppose an entire motion covering all the discovery materials in a case to which they are not a party.

Lastly, we believe that plaintiffs have effectively waived their right to seek wholesale declassification. The parties voluntarily entered into PTO 35 in order to facilitate discovery. See note 10 supra. They agreed upon a method by which to challenge confidentiality designations. They also agreed that the information produced would be used only by “qualified” individuals and would be used only in the preparation for trial “and shall not be used for any other purpose whatsoever.” Plaintiffs are now attempting to abrogate that agreement. PTO 35 did not contemplate such an unwieldy approach. Rather, PTO 35 instituted a more particularized approach for objections. All parties have relied on these provisions in producing documents, and wholesale declassification would undermine their justified expectations. Plaintiffs cannot now attempt to undo what they have willingly wrought; having made their bed, they must sleep in it.

We note that Judge Frankel reached a similar conclusion in GAF Corp. v. Eastman Kodak Co., 415 F.Supp. 129 (S.D.N.Y.1976). In that antitrust case, the parties produced tens of thousands of documents under an agreed order of confidentiality, providing that all documents produced in discovery would be used solely for the purpose of that litigation. GAF subsequently sought permission to turn fifty-two documents over to the Antitrust Division of the Department of Justice. Judge Frankel denied their request, concluding that the parties’ “shared and explicit assumption” that discovery in the case was for purposes of that case alone went “a long way toward denial of GAF’s request without more.” Id. at 132.

Inasmuch as there is a paucity of authority on this point, 48 we rest our decision essentially upon our own analysis. We note, however, that our conclusion is supported by the pithy and forceful opinion of Judge Russell Smith in In Re Coordinated Pretrial Proceedings in Western Liquid Asphalt Cases, 18 F.R.Serv.2d 1251 (N.D.Cal.1974). 49

Although we deny plaintiffs’ motion for wholesale declassification as such, we will treat the motion as one seeking declassification of the critical categories of documents in the case. The task of deciding whether certain categories of documents should be released requires us next to consider the matter of the common law access rights to judicial records.

III. Common Law Access Rights to Judicial Records 50

A. The General Nature of the Right

It is beyond dispute that in this country judicial proceedings and records are presumptively open to the public. 51 The public’s right of access was recognized at common law before the Constitution was adopted. United States v. Criden, 648 F.2d 814, 819 (3d Cir. 1981). Judicial openness has two components: access to trials themselves, which for criminal trials the Supreme Court has found is protected by the First and Fourteenth Amendments, Richmond Newspapers, Inc. v. Virginia, 448 U.S. 555, 100 S.Ct. 2814, 65 L.Ed.2d 973 (1980); and access to judicial records for inspection and copying, a privilege which the Court explicitly recognized in Nixon v. Warner Communications, Inc., 435 U.S. 589, 597-99, 98 S.Ct. 1306, 1311-13, 55 L.Ed.2d 570 (1978). The right to inspect and copy, part of a more general citizens’ right to review public records, United States v. Mitchell, 551 F.2d 1252, 1257-58 (D.C.Cir.1976), rev’d on other grounds sub nom. Nixon v. Warner Communications, Inc., is the more relevant to our determination of the issues in this case. However, because the Third Circuit has concluded that identical interests support both the public’s right to inspect and copy and the public’s right of access to criminal trials, see United States v. Criden, supra, 648 F.2d at 820, we will also canvass the trial access cases.

The right to inspect and copy has been discussed in recent years by the Supreme Court and several courts of appeals, including the Third Circuit, though its precise contours have not been determined.' In Nixon v. Warner Communications, Inc., which was decided on statutory grounds, the Supreme Court did not explore in depth the underlying bases of the right. The District of Columbia Circuit declared in that case, however, that “this common law right is not some arcane relic of ancient English law. To the contrary, the right is fundamental to a democratic state.” United States v. Mitchell, supra, 551 F.2d at 1258. The court explained that the right of inspection, like the First Amendment, assures a well-informed public opinion, permits public monitoring of the courts, and promotes confidence in the fairness and justice of the court system. 52 In United States v. Criden, the Third Circuit noted that this right has been “justified on the ground of the public’s right to know, .which encompasses public documents generally, and the public’s right to open courts, which has a particular applicability to judicial records.” 648 F.2d at 819.

The most recent Supreme Court decision to consider the public’s right of access is Richmond Newspapers, Inc. v. Virginia, which held that under the First Amendment the public has access rights to criminal trials that are independent of the accused’s Sixth Amendment right to a public trial. The Chief Justice’s plurality .opinion reviewed the venerable history of public trials, concluding that judicial openness serves a “significant community therapeutic value,” 448 U.S. at 570, 100 S.Ct. at 2824, and that “openness, fairness, and the perception of fairness” are directly linked to one another. Id. Thus, he explained, “[pjeople in an open society do not demand infallibility from their institutions, but it is difficult for them to accept what they are prohibited from observing.” Id. at 572, 100 S.Ct. at 2825. The Chief Justice’s thorough review of the tradition of and public interest in open judicial proceedings led him to conclude that “the right to attend criminal trials is implicit” in the First Amendment guarantees of free speech, free press, and the right of assembly. 53 Though the plurality opinion by the Chief Justice considered only the public’s right to attend criminal trials, he noted that “historically both civil and criminal trials have been presumptively open.” Id. at 580 n.17, 100 S.Ct. at 2829 n.17. We have no doubt that the interests he enumerates are equally applicable to civil trials, whether or not access is deemed to be guaranteed by the First Amendment.

To the same effect as Richmond Newspapers is the opinion of Chief Judge Seitz for the Third Circuit in United States v. Cianfrani, 573 F.2d 835 (3d Cir. 1978). 54 Chief Judge -Seitz stated there that “our law strongly favors the presumption that all adjudicative proceedings are open to the public.” Id. at 847. 55

As has been explained, a right of access for inspection and copying attaches to public records, including court records. But the cases do not support a right of access to materials that are involved in litigation but never come into the possession of the court. Moreover, the cases are unclear as to the right of access to materials that are in the temporary possession of the court for purposes of evidentiary rulings, but which are not received into evidence. Finally, the law is simply not yet developed with respect to documents that play other roles in court proceedings. Therefore, we must decide what documents comprise the judicial record. We turn now to that task.

B. The Extent to Which Access Rights Attach to Judicial Records

1. Introduction

The types of materials which comprise the judicial record and to which the common law right to inspect and copy therefore attaches have not been thoroughly catalogued. The right undoubtedly attaches to all materials that are filed with the clerk of court, unless filed under seal pursuant to court order, because they are so clearly records within the meaning of the doctrine. These materials include the complete transcripts of testimony, legal argument, and other proceedings in open court. Federal law requires that open district court proceedings be reported and that the reporter’s “original notes or other original records and the copy of the transcript in the office of the clerk be open during office hours' to inspection by any person.” 28 U.S.C. § 753(b) (1976). Thus, portions of documents that are read into the record in an open proceeding enter the public domain, without regard to whether the documents were originally filed on the .public record, were filed under seal, or were not filed with the clerk. In addition, the public would have access in the ordinary case to “all papers required after the complaint to be served upon a party,” which must be filed under Rule 5(d). 56 Though the availability for inspection and duplication 57 of trial exhibits and materials admitted into evidence has been questioned, such materials have been determined to be judicial records available to the public. See, e.g., United States v. Criden, supra, 648 F.2d at 822; United States v. Mitchell, supra, 551 F.2d at 1259.

The case law provides us with little guidance, however, in determining what materials other than public filings, which include transcripts of hearings and exhibits received in evidence, are part of the judicial record. We think that whether various kinds of materials are part of the judicial record is a question that must be answered for each in light of the purposes served by the common law right to inspect and copy public records. Bearing in mind that access rights exist to promote knowledge of and attention to the performance of the courts, for the benefit of society as a whole, we turn to that task.

2. Discovery Materials

Writing for the court in Wilk v. American Medical Ass’n, 635 F.2d 1295, 1299 n.7 (7th Cir. 1980), Judge Wisdom observed that:

[U]nless and until introduced into evidence, 58 the raw fruits of discovery are not in the possession of a court. If the purpose of the common law right of access is to check judicial abuses, see United States v. Mitchell, 551 F.2d 1252, 1257 — 58 (D.C.Cir.1976), rev’d sub nom. Nixon v. Warner Communications, Inc., 435 U.S. 589 [98 S.Ct. 1306, 55 L.Ed.2d 570] (1978), then that right should only extend to materials upon which a judicial decision is based.

We agree with Judge Wisdom that the public has no common law right to inspect materials that are produced in discovery but are not placed in the custody of the court. 59 We have found no decisions that reach a contrary conclusion under the common law. 60

3. Material that is the Subject of an Evidentiary Ruling

We have mentioned previously that trial exhibits are part of the judicial record to which access rights attach. In this section, we consider the status of materials that are proffered to the court in an open hearing but are ruled inadmissible.

In United States v. Gurney, 558 F.2d 1202, 1210 (5th Cir. 1977), cert. denied sub nom. Miami Herald Publishing Co. v. Krentzman, 435 U.S. 968, 98 S.Ct. 1606, 56 L.Ed.2d 59 (1978), the Fifth Circuit Court of Appeals concluded that various documents and exhibits presented in a criminal trial for identification but “not yet admitted into evidence” had not become part of the public record. In a decision that may be reconciled with Gurney because of its different emphasis, the District of Columbia Circuit Court of Appeals suggested that when documents are produced to the court with a request for a ruling, they become part of the record whether they are or are not technically admissible into evidence. United States v. Hubbard, 650 F.2d 293, 299 n.11 (D.C.Cir.1980). 61

We agree with the view of the District of Columbia Circuit. If the public’s right to inspect and copy serves educative and monitoring functions, it is illogical to define the public record so as to preclude review by the public of evidentiary rulings, especially in situations where, as here, evidentiary rulings are critical to the ultimate disposition of the case. Our conclusion is buttressed by Local Rule 39(d), which provides that:

All exhibits received in evidence, or offered and rejected, upon the hearing of any cause or motion, shall be delivered to the Clerk, who shall keep the same in custody, unless otherwise ordered by the court.

We also think that Rules 10(a) and (b)(2) of the Federal Rules of Appellate Procedure, defining the district record for purposes of appeal, are instructive. 62 The record on appeal generally includes the papers and exhibits filed, the transcripts, and a certified copy of the docket; in addition, an appellant who will argue that some aspect of the trial court’s ruling is not supported by the evidence must include in the record “a transcript of all evidence relevant to such finding or conclusion.”

In sum, we think that all materials that are the subject of an evidentiary ruling by the court, whether or not found admissible, are part of the record for purposes of the public’s right to inspect and copy. 63

4. Materials Referred to at a Hearing

We will next discuss whether access rights attach to materials to which public reference is made at a hearing but which are neither read into the record nor offered into evidence. This question is a difficult one, and we receive very little guidance in its resolution from reported cases. At the same time, we must resolve this question to decide the declassification motion. A very significant portion of the documents that are the subject of the motion were dis cussed briefly, summarized, or otherwise mentioned during the in limine hearing or during the summary judgment hearing. It is clear, of course, that whatever information from or about those documents appears in the public record filings is already in the public domain.

The case that we have found most relevant to this inquiry is United State s v. Hubbard, supra. The Hubbard court reviewed a district court order unsealing a large number of papers that had been seized from a third party nondefendant to a federal criminal prosecution. The papers had been produced in court only under seal and only in a hearing on defendants’ motion to suppress the documentary evidence because it was allegedly illegally seized. Relatively few of these documents were ruled upon or referred to either in the hearing or in the trial judge’s decision on the motion. The District of Columbia Circuit nonetheless had no difficulty concluding that all of the documents were part of the record in the case, because the full set of papers was urged on the judge in the suppression hearing. 650 F.2d at 299-300. In other words, access rights might attach to the papers even though only very general reference was made to them in the hearing. The court indicated that the brevity or generality of the reference was a factor to be weighed in determining whether the papers should properly remain sealed despite their inclusion in the record. 64 See id. at 314-17; supra note 61.

We agree with the District of Columbia Circuit that documents that are referred to at a hearing do become part of the record, 65 and we further agree that the nature of the reference may support the case for nondisclosure of sealed record material if the reference is extremely general or noninformational and if countervailing interests are shown. 66 General references to material alleged to contain sensitive commercial information are more likely to support continued protection, for example, than detailed references would be. If a significant amount of the allegedly sensitive data is already in the public domain, because included in the publicly filed transcript of a proceeding, then release of the documentary material containing such data will cause relatively little further injury. If, on the other hand, almost none of the data has been made public, releasing the documentary material could inflict serious competitive injury. 67

Notwithstanding the foregoing qualifications, we believe that a rule that arbitrarily excluded from the judicial record documents that were not presented to the court individually and in their entirety would not serve the important purposes of the common law right to inspect and copy. We see no reason to assume that such documentary materials are never pertinent to understanding a court’s decisions or evaluating judicial performance. Moreover, in some cases there might be no interests that could rebut even a weak presumption of public access.

5. Records Filed Under Seal that are the Ultimate Subject of a Dispositive Ruling

The final category of documents that we consider is that of record materials on which the court relies in making a ruling and which the court discusses in a published opinion. If such documents are themselves filed, the fact of filing, of course, triggers potential public access rights. Even if such documents are not filed, however, special reasons appear for their disclosure.

We think that the courts have an obligation to explain their decisions that is correlative to the public’s common law access rights. 68 The core function of the right to inspect and copy judicial records can only be served by the release in full of documents that are the basis of a dispositive ruling. Unlike the documents at issue in Hubbard, the release of which the District of Columbia Circuit found would not significantly increase public understanding of that case, see 650 F.2d at 317 — 18, the disclosure of documents relied on by the court often may be critical to a proper evaluation of the decision. As anyone familiar with this litigation realizes, that is the case here.

C. Countervailing Interests: The Balancing Test

The common law right to inspect and copy is of great importance, but it is not absolute. The tradition of public access has as its corollary the notion that courts must be free to supervise their records and files and to limit access to them when necessary to prevent improper use.

For example, the common-law right of inspection has bowed before the power of a court to insure that its records are not “used to gratify private spite or promote public scandal” through the publication of “the painful and sometimes disgusting details of a divorce case.” Similarly, courts have refused to permit their files to serve as reservoirs of libelous statements for press consumption, or as sources of business information that might harm a litigant’s competitive standing.

Nixon v. Warner Communications, Inc., supra, 435 U.S. at 598, 98 S.Ct. at 1312 (citations omitted). Judicial proceedings and records may be closed in part or in full to the public in order to protect private interests, including proprietary interests in trade secrets and other commercial information. See, e.g., Stamicarbon, N. V. v. American Cyanimid Co., 506 F.2d 532, 539-40 (2d Cir. 1974); Note, All Courts Shall Be Open: The Public’s Right to View Judicial Proceedings and Records, 52 Temp.L.Q. 311, 335 & n.214 (1979) (citing cases).

To decide whether the interests that favor the sealing of records are weighty enough to permit closure, the court must balance these interests against the presumption of public access. The strength to be accorded the presumption of access, however, varies with the nature of the materials at issue. See Nixon v. Warner Communications, Inc., supra, 435 U.S. at 602, 98 S.Ct. at 1314; United States v. Criden, supra, 648 F.2d at 822-23. In this section, we will review factors that courts have considered pertinent to an inquiry whether judicial records should be sealed. Also, we will discuss the strength of the presumption of access that attaches to different kinds of record materials. 69

United States v. Criden involved an application by members of the press 70 to copy video and audio tapes that had been received as exhibits in a public criminal trial. 71 Complete transcripts of the tapes were released to members of the press and the public, but the request to copy was denied by the district court. On appeal, the Third Circuit speaking through Judge Sloviter, detailed the interests that appeared for and against release. The preeminent factor favoring release was found to be the public’s right to inspect and copy judicial records. That right was “buttressed by the significant interest of the public in observation, participation, and comment on the trial events.” 72 648 F.2d at 823.

Against these factors, .the court weighed the special nature of audio and videotape evidence, to determine whether the likely “replication of the trial experience” militated against release. To the contrary, the court thought that rebroadcast would serve an educative function, and rejected the argument that rebroadcast would enhance the defendants’ punishment. Id. at 824-25. The most important argument for nondisclosure in the eyes of the court was the risk that defendants’ rights to a fair trial would be infringed, but the court emphasized that the trial court should have considered less restrictive means to accommodate these rights. Id. at 827, citing Chandler v. Florida, 449 U.S. 560, 101 S.Ct. 802, 66 L.Ed.2d 740 (1981).

Finally, the court addressed the district court’s concern that dissemination of the tapes might invade the privacy of nondefendants, concluding that “the district court has some area of discretion in which to balance the strong public interest favoring access against legitimate privacy concerns of third parties.” 648 F.2d at 829. Even here, the court of appeals required that the district court accommodate nonparty interests in the least restrictive way available by excising specific portions of the tapes. The case was remanded for consideration of possible privacy interests.

From the Criden opinion, we derive several principles to guide our decision in this very different case. First, the presumption of access to evidentiary materials is strong. We think therefore that the presumption applicable to materials publicly filed of record must be very strong indeed. These court papers, transcripts, and other materials are a principal focus of the right to inspect and copy. Second, where the balance of interests favors the denial of access to any materials, any infringement of public access rights must be no greater than is necessary to afford the necessary protection. 73

In United States v. Hubbard, the District of Columbia Circuit weighed the factors that favored nondisclosure of materials seized from nondefendants and referred to in only a general way during a suppression hearing. Under the circumstances of that case, the court thought that continued protection might be warranted. 74 The most important consideration leading to this conclusion was the purpose for which the documents at issue were introduced: solely to demonstrate the alleged Fourth Amendment infirmity of their seizure. They were ruled irrelevant to the underlying prosecution and not described or relied, on in the district court’s ruling on the suppression motion. The court noted that “it would be ironic indeed if one who contests the unlawfulness of a search and seizure were always required to acquiesce in a substantial invasion of those interests simply to vindicate them.” 650 F.2d at 321 (footnote omitted). In addition, the Hubbard record revealed that the nondefendant Church of Scientology from which the documents were taken had property and privacy interests that were strong enough to be recognized. By contrast, the incremental public benefit that would be achieved by disclosure was small:

The public in this case had access, inter alia, to the courtroom proceedings on the motion to suppress, to the memoranda filed by the parties in connection with that motion, to the trial judge’s memorandum decision on the suppression motion, to the trial judge’s memorandum decision on the negotiated disposition, to the stipulated record which was the basis for the defendants’ convictions and to the actual “trial” of the criminal charges of which the defendants were convicted. None of the documents at issue here was either used in the examination of witnesses during the protracted public hearing on the suppression motion or specifically referred to in the trial judge’s public decision on the motion to suppress or included as part of the publicly available stipulated record on which the defendants’ criminal convictions were had.

Id. at 317-18.

The court also found it significant that until the unsealing order appealed from was entered, the public had not had access to the documents. 75

The Tenth Circuit considered the propriety of retaining under seal appellate briefs that discussed and quoted from some of several hundred documents that were subject to a district court protective order in Crystal Grower's Corp. v. Dobbins, 616 F.2d 458 (10th Cir. 1980). Appended to the briefs was a copy of the trial court’s sealed opinion that relied on the documents. That court had ordered that the documents be filed under seal because the producing party asserted attorney-client privilege and work-product immunity in them. Oral argument before the court of appeals was open to the public, but the content of individual documents was not discussed. 76 Weighing the interests of the public against those of the parties, 77 the Tenth Circuit found that disclosure was not warranted. First, the court noted that while the general public interest in understanding and monitoring judicial proceedings favored unsealing, the attorney-client privilege and work-product immunity doctrines incorporated a specific public interest. That interest was the encouragement both of full and unimpeded communication between clients and their counsel and of complete preparation by attorneys. In addition, the court found that one party had a substantial interest in nondisclosure, which is not unlike that recognized in Hubbard : release would vitiate the claims of privilege and immunity without a hearing on the merits. Id. at 460-61.

United States v. Hubbard and Crystal Grower’s Corp. v. Dobbins are among the few decisions we have found that have balanced the interests favoring nondisclosure against the presumption of public access and found the former to be the weightier. 78 Both cases presented unusual factual situations in which the parties opposed to disclosure would suffer very serious injuries as a consequence of release, in which much of the information contained in the documents at issue had not entered the public domain, and in which the public interest in access to those documents was found to be particularly weak. The District of Columbia Circuit, for example, stressed that almost all materials relating to the Hubbard prosecution other than the allegedly illegally seized documents were accessible to the public, and the Tenth Circuit noted in Crystal Grower’s that no petitioner had appeared to assert a particular need for access, 616 F.2d at 462 n.2. From these cases, we conclude that the presumption of access is weak when the purposes of the right to inspect and copy are served without affording access to materials that were discussed or referred to only in a general way.

Our consideration of United States v. Criden, United States v. Hubbard, and Crystal Grower’s Corp. v. Dobbins suggests that access rights must be evaluated on a case-by-case basis. The more fully materials are included in the public case record, the stronger the presumption of access. Criden implies that the presumption of access to record filings should be very strong, and holds that the presumption of access to evidentiary materials should be strong. Along with the District of Columbia and Tenth Circuits we think that the presumption of access to materials merely referred to in a general way is relatively weak, at least if the general purposes of the right to inspect and copy are well served absent release of those materials. When the degree of disclosure on the record of the materials at issue is greater than mere reference but less than full incorporation, we think that the proper weight can be assigned to the presumption of access only after consideration of the amount and nature of material already on the public record. 79

D. Access Rights to the Critical Documents at Issue

I. Introduction

Before we proceed to strike the balance that determines whether the common law right of access to public records requires release of certain types of documents, we think it appropriate to enumerate the interests for and against disclosure that appear in this case. These interests will be weighed differently with respect to each category, but can be usefully discussed as a group.

The factors that favor disclosure include the general presumption of public access, as described in its various aspects in the preceding segment of this opinion, and a particular public interest in antitrust proceedings. Private plaintiffs bringing damages actions under the Clayton Act are “private attorneys general,” whose successful lawsuits vindicate societal interests as well as their own. See, eg., Hawaii v. Standard Oil Co., 405 U.S. 251, 262, 92 S.Ct. 885, 891, 31 L.Ed.2d 184 (1972); Perma Life Mufflers, Inc. v. International Parts Corp., 392 U.S. 134,139, 88 S.Ct. 1981, 1984, 20 L.Ed.2d 982 (1968).

There are also public interests favoring nondisclosure in this case. First, there can be no doubt that society in general is interested in the protection of trade secrets and other valuable commercial information. That interest is recognized, for example, in Rule 26(c)(7), in our copyright, trademark, and patent statutes, and in the common law of business torts. These policies recognize a coincidence of public and private interests. Second, we think that orderly management of complex litigation is in the public interest. Enormous cases like this one cannot be fairly and expeditiously adjudicated unless parties are assured that their legitimate interests will be protected. Often that assurance cannot be given unless parties are permitted to rely on guarantees of confidentiality.

With these interests in mind, we turn to the documents at issue.

2. The Document Submission Sheets

The question of access rights to the DSS’s need not detain us long. We have noted that the DSS’s were introduced, discussed, and considered at great length in a pretrial evidentiary hearing which we have concluded was open to the public. 80 We have also seen that the DSS’s were the subject matter of evidentiary rulings which are set forth in three lengthy opinions filed on the public record and disseminated through legal publications.

We find further support for our decision on the question of access to the DSS’s in the fact that a large number of the DSS’s are quoted in full in the FPS, which will be declassified. Moreover, large portions of the JFTC record, which comprises a significant portion of the DSS’s, were open to the public in Japan. 81 We need not rely upon these additional factors, however. We deal here with materials that were proffered to the court and made the subject not only of an evidentiary hearing, but of an evidentiary ruling. We have concluded as a matter of law that such materials, whether or not found admissible, are part of the record for purposes of the public’s right to inspect and copy, 82 and that the record of the evidentiary hearing is public.

Although we are conscious of our obligation to balance countervailing interests, the presumption of access to the DSS’s is so very strong that the balance is not a close one. We have considered the defendants’ interests in protecting their valuable commercial information. The DSS’s, however, contain relatively small amounts of such information with the exception of Dr. De-Podwin’s report, see supra page 883. That report extrapolates from the distilled raw economic data, presenting its conclusions in a form that is useful for argumentative purposes but not useful for competitive purposes. Much of the economic data is stale, moreover, and it would take a Herculean effort for a competitor to put it to use. The information relative to dealings between the Japanese manufacturing defendants with private label and original equipment manufacturer customers is so spotty and incomplete that there is little danger of competitive injury from its disclosure.

3. The Final Pretrial Statement

We have explained that though the FPS was filed under seal, it was nonetheless the principal focus of the summary judgment hearing, from which the public was never excluded. The critical question before us in connection with the summary judgment proceedings was whether there was anything in the FPS that created a genuine issue of material fact as to the existence of a Sherman Act conspiracy among the defendants. Our lengthy opinion granting summary judgment focused on this question. As we have explained, the sealed documents quoted or cited in the FPS were a constant subject of colloquy at the hearings.

We have concluded that courts have an obligation to explain their decisions that is correlative to the public’s common law access rights, and that the core function of the right to inspect and copy judicial records can only be served by the release in full of documents that are the basis of a dispositive ruling. Because our summary judgment opinion cannot be understood or evaluated without the release in full of the FPS, the FPS must be declassified. 83

We do not reach this conclusion without consideration of countervailing interests. We have determined, however, that the balance here favors disclosure because of the strength of the presumption of access and the public’s interest in antitrust actions, compared to the lesser force of the interests favoring confidentiality. Our decision that the FPS must be public may aid public supervision of complex cases like this one. To be sure, there is a considerable amount of sensitive commercial information contained in the FPS, including information as to the defendants’ prices at various levels of distribution, and the prices charged to various customers. But this information is not arrayed usefully in the body of the FPS, as it is in the computer generated model-by-model matchups and price comparisons and defendants’ comprehensive answers to interrogatories. Rather, it is arrayed for use by a lawyer constructing an argument, and it is spread across thousands and thousands of the FPS. It thus would be enormously burdensome to extrapolate any useful information therefrom. In short, the balance plainly favors disclosure. 84

4. Documents Referenced in the Summary Judgment Hearings and in the Final Pretrial Statement

The references at the summary judgment hearing to documents alleged to contain sensitive commercial information were general, not detailed. Although some of this allegedly sensitive data is already in the public domain because it is included in the publicly filed transcript, or is contained in the DSS’s or the FPS which have been declassified, we view this part of plaintiffs’ motion in its “wholesale” aspect. Plaintiffs have not canvassed the summary judgment hearing record and tendered to us a list of documents referred to therein, nor have we considered it an appropriate expenditure of judicial time for us to do so. There may well have been documents referred to at the hearing which contain sensitive commercial information, but which are not already in the public domain. To that extent, the countervailing interests in protection of this material would be strong.

We have concluded in our discussion of the applicable law that although documents that are referred to at a hearing become part of the record, the nature of the reference may support nondisclosure of record materials if the reference is extremely general or noninformational and if countervailing interests are shown.

We have also determined that the presumption of access to materials referred to in a general way is very weak, and that the presumption is weaker still when the general purposes of the right to inspect and copy are well served absent release of those materials. These conditions are all present here, in part because of our other declassification orders. Under these circumstances, we will deny access to documents merely referenced in the summary judgment hearing.

We need not labor the question whether the motion before us requires us to classify documents that are referenced in the FPS. We have already concluded, see supra note 65, that documents that are referenced in papers filed on the public record are not part of that record for access purposes.

5. The “Raw Economic Data” Contained in Sealed Discovery Responses and the Final Pretrial Statement Appendices

We have explained above how the defendants have filed under seal extensive raw economic data, some of which has been attached by the plaintiffs as appendices to the FPS. See note 84 supra. This material includes such matters as the prices charged by defendants for various models of television receivers at different levels of distribution, and documents reflecting certain dealings with specific customers. This material does not fall into the categories we have discussed in Part III-B of this opinion: it is not material that is the subject of an evidentiary hearing; it is not material referred to at a hearing; and it is not material that was the ultimate subject of a dispositive ruling. It is, because it has been filed in the record of the court, material subject to an access right, but, as we have seen, that access right is not absolute. The presumption of access to this economic data is not strong. The public’s general educative and monitoring interests, as well as its specific interest in antitrust and international trade matters, are essentially satisfied by our declassification of the FPS and of the DSS’s, especially the expert witness reports.

On the other hand, the interests favoring nondisclosure are strong. The raw economic data obviously contains sensitive commercial information. Although the information is old, we are persuaded by the submissions by Mr. Miyamoto and Mr. Yamagishi, see notes 21 and 22 supra, that significant harm could come to the defendants from disclosure of this information in its raw form. Accordingly, we will deny access to the raw economic data. 85

Our determination that the public’s access rights do not, on balance, require the release of some categories of documents does not dispose of plaintiffs’ motion. Because plaintiffs urge that their First Amendment rights are infringed by continued nondisclosure, we must decide whether the Constitution requires more than does the common law right of access.

IV. First Amendment Interests in Judicial Records

A. Introduction

The First Amendment’s Protection of the freedoms of speech and press from unjustifiable governmental restraints includes protection from judicial orders. See, e.g., Nebraska Press Association v. Stuart, 427 U.S. 539, 96 S.Ct. 2791, 49 L.Ed.2d 683 (1976); Craig v. Harney, 331 U.S. 367, 67 S.Ct. 1249, 91 L.Ed. 1546 (1947). Because the First Amendment unquestionably protects the dissemination of information about court proceedings, we must consider whether Rule 26(c) protective orders that are imposed on a showing of good cause are or may be constitutionally invalid prior restraints of speech. See, e.g., In re San Juan Star Co., 662 F.2d 108 (1st Cir. 1981); In re Halkin, supra.

In addition to inquiring into litigants’ First Amendment interests in discovery materials, we must decide whether the First Amendment affords the public access rights to judicial records. The Supreme Court has found that the Constitution does guarantee access to criminal trials under some circumstances. Richmond Newspapers, Inc. v. Virginia, supra. In United States v. Criden, supra, 648 F.2d at 821 n.6, the Third Circuit noted that the question whether there is a correlative constitutional guarantee of access to court records was left open by Richmond Newspapers. The Third Circuit found it unnecessary to answer the question in Criden, which was decided on nonconstitutional grounds.

We will address each of these First Amendment questions in turn. With respect to Rule 26(c) protective orders, we conclude that litigants’ First Amendment interests in the dissemination of discovery materials are adequately protected by the balancing of interests for and against nondisclosure and drafting such orders in the least restrictive manner possible. With respect to the question whether the common law right to inspect and copy has a constitutional dimension, we conclude that it does not.

B. First Amendment Rights in Discovery Materials

Plaintiffs assert that their First Amendment rights to disseminate information in their possession is violated by our continued enforcement of PTO 35. Arguing that Rule 26(c) protective orders are prior restraints of speech, they contend that PTO 35 cannot survive analysis in light of “ ‘[t]he heavy presumption against . . . constitutional validity’ that an order restraining publication bears,” Nebraska Press Association v. Stuart, supra, 427 U.S. at 545, 96 S.Ct. at 2796, quoting New York Times Co. v. United States, 403 U.S. 713, 714, 91 S.Ct. 2140, 2141, 29 L.Ed.2d 822 (1971).

The Third Circuit has not decided whether the dissemination of information contained in discovery materials is entitled to First Amendment protection. 86 The best known opinion that analyzes protective orders under the First Amendment, and concludes that they must be viewed as prior restraints of speech, is that of the District of Columbia Circuit Court of Appeals in In re Halkin, supra. 87 In a very recent opinion, In re San Juan Star Co., supra, the First Circuit decided that Rule 26(c) protective orders implicate “significant but limited” First Amendment interests. Id., at 114. Chief Judge Coffin for the First Circuit announced a less exacting test of constitutionality than did the District of Columbia Circuit, one that we think better takes account of the special nature of discovery materials. We will review both decisions and then explain why we find San Juan Star the more persuasive.

In In re Halkin, the court reviewed a district court order that restrained the parties and their counsel from discussing or disclosing publicly any “documents and information” obtained from certain federal agencies through discovery. 88 The court found that the order “pose[d] many of the dangers of a prior restraint,” so that its impact on constitutionally protected expression had to be scrutinized closely, 598 F.2d at 186, and then proceeded to examine the First Amendment interests of the litigants that were affected by the district court order. Because the right to disseminate information does not depend on the method of its acquisition, cf. New York Times Co. v. United States, 403 U.S. at 745, 91 S.Ct. at 2157 (refusing to restrain publication of stolen government documents), and participation in civil litigation cannot by itself constitute a waiver of First Amendment rights, 89 cf. Curtis Publishing Co. v. Butts, 388 U.S. 130, 145, 87 S.Ct. 1975, 1986, 18 L.Ed.2d 1094 (1967) (declining to find waiver “in circumstances which fall short of being clear and compelling”), the court determined that discussion of information obtained through discovery was protected expression. Therefore, the Halkin court concluded, any protective order that seriously restrains expression must satisfy a three-part test: 90 “The harm posed by dissemination must be substantial and serious; the restraining order must be narrowly drawn and precise; and there must be no alternative means of protecting the public interest that intrudes less directly on expression.” 598 F.2d at 191 (footnotes omitted).

Judge Wilkey, dissenting, found the majority’s prior restraint analysis of protective orders misconceived. He argued that the good cause standard of Rule 26(c) amply protected the legitimate interests of litigants. Indeed, he thought the constitutional propriety of the order issued by the district court in Halkin “plain.” 598 F.2d at 203. He explained:

The First Amendment interests of litigants in the promulgation of materials exacted from another party through the compulsory processes of the courts are much more limited and of a fundamentally different character from the First Amendment interests of litigants and non-litigants in the public communication of other information concerning judicial proceedings. This is because litigants who wish to disseminate discovery materials have gained access to such materials — access which they would not ordinarily have — through a statutory system that expressly reserves to the courts the power to attach restrictions on the use of such materials.

598 F.2d at 206 (Wilkey, J., dissenting). Judge Wilkey concluded that the district court’s power under the Rules of Civil Procedure to limit or condition the use of information obtained in discovery is necessary to “overall fairness” because of the broad scope of discovery permitted by the Rules.

In San Juan Star, the First Circuit reviewed a district court order that barred counsel in a civil rights action from disseminating to the parties or the public any information obtained through deposition testimony. 91 That court concluded that Rule 26(c) protective orders implicate First Amendment interests, but that the interests of litigants, their attorneys, and others in discovery materials are more limited than their interests in disseminating other kinds of information. 92 Because discovery rules are liberal, much of the information obtained thereby may never be used at trial; it may be inadmissible because it is irrelevant or unduly prejudicial. 93 For that reason, the court held that raw deposition testimony is not information entitled to full First Amendment protection because its disclosure would not serve to inform “civic and political discussion,” to permit monitoring of the judicial system, or to implement values of judicial openness. At 115. The San Juan Star court was unwilling to deprive discovery materials of all constitutional protection, however, since the First Amendment expresses a concern that “the government not lightly engage in any restraints on communication, particularly when the order is issued prior to the expression taking place.” Id. The important role played by protective orders in facilitating discovery and expediting litigation was also noted. 94

In light of these considerations, the court held that the proper standard for determining the constitutional validity of a protective order is a less stringent than usual application of the test applied to prior restraints. Thus the court would examine the nature of the threatened harm, the effectiveness of the order, the availability of less restrictive alternatives, and the narrowness of the order decreed. The magnitude and imminence of the threatened harm would be evaluated on a “sliding scale: as the potential harm grows more grave, the imminence necessary is reduced.” 95 At 116.

San Juan Star announces a test for protective orders that successfully takes account of First Amendment concerns while preserving the trial court’s broad discretion to impose protective orders for good cause. We think that a satisfactory alternative description of the proper standard is that it is a balancing test. The balance of interests may justify a prior restraint when “the restraint is only reasonably incidental to the achievement of another valid governmental purpose.” New Jersey State Lottery Comm’n v. United States, 491 F.2d 219, 222 (3d Cir. 1974) (en banc), vacated on other grounds, 420 U.S. 371, 95 S.Ct. 941, 43 L.Ed.2d 260 (1975); see United States v. O’Brien, 391 U.S. 367, 376, 88 S.Ct. 1673, 1678, 20 L.Ed.2d 672 (1968).

The overriding purpose of a protective order is to facilitate the communication of information between litigants. 96 To accom plish that end, it may be necessary to limit speech by parties and their counsel outside the court. Several general governmental interests support the imposition and enforcement of such orders. First, the courts must preserve the integrity of the discovery system and protect litigants from “annoyance, embarrassment, oppression, or undue burden or expense,” Rule 26(e), and must guard against abuse of their own processes. Second, the courts must be able to manage successfully large and complex cases. 97 Third, courts have a general responsibility to do justice. Litigants in many cases cannot be guaranteed a fair trial of their claims and defenses if necessary documents and materials are not produced; at the same time, the court must seek to protect from unwarranted harm parties whose rights may ultimately be vindicated at trial. See In re San Juan Star Co., supra, at 115.

To apply a balancing test, of course, the particularized interests of the litigants for and against disclosure must be weighed. Litigants do have a general interest in publicizing information that they possess, however acquired. We think it worth emphasizing, however, that even when a protective order is entered, litigants have access to discovery materials for all the purposes for which they are legitimately acquired through court processes.

In sum, we think that the proper standard for deciding whether plaintiffs’ First Amendment interests would be violated by our deciding not to declassify certain documents is not a “heavy presumption” of unconstitutionality, but a balancing test. The court must evaluate the magnitude and imminence of the threatened harm from disclosure, including the particularized interests of the litigants against disclosure and the general government interests enumerated above, the narrowness of the proposed limiting order, 98 and the availability of less restrictive alternatives. We cannot, of course, delineate precisely the application of the test except on a case-by-case basis. 99

C. Relationship Between the First Amendment and the Common Law Right to Inspect and Copy

In United States v. Criden, supra, the Third Circuit left open the question whether the First Amendment incorporates a right of the public to access to judicial records, but noted that Richmond Newspapers arguably “could be viewed as supporting a right of the public to access to [trial exhibits] through the medium of the broadcasters.” 648 F.2d at 821 n.6. We must now address that issue because we have concluded that the balancing of interests will sometimes defeat the common law right of access. We conclude, however, that a constitutional right of access to judicial records does not exist.

At the outset, we emphasize that we address only an asserted constitutional right of access to materials that are part of the case record, but have not yet entered the public domain. The principle that the First Amendment protects the communication of information contained in public court records is well established. E.g., Cox Broadcasting Corp. v. Cohn, 420 U.S. 469, 495, 95 S.Ct. 1029, 1046, 48 L.Ed.2d 328 (1975).

In Nixon v. Warner Communications, Inc., supra, the Supreme Court rejected the contention that the First Amendment guarantees the press access, for the purpose of inspection and copying, to materials and exhibits used in open judicial proceedings. The court found a distinction of constitutional dimension between access to the information contained in trial exhibits and physical access to the exhibits themselves, explaining that in that case:

There simply were no restrictions upon press access to, or publication of any information in the public domain. Indeed, the press — including reporters of the electronic media — was permitted to listen to the tapes and report on what was heard. Reporters also were furnished transcripts of the tapes, which they were free to comment upon and publish. The contents of the tapes were given wide publicity by all elements of the media. There is no question of a truncated flow of information to the public.

435 U.S. at 609, 98 S.Ct. at 1318.

Warner Communications rejected the possession by the press of a First Amendment right to inspect and copy, and did not consider possible access rights of the general public. We think nonetheless that the case is properly read to hold that the public’s right to inspect and copy is not of constitutional stature. We can conceive of no reason why the public’s First Amendment rights in this area should be greater than those of the press. 100

We do not think that the First Amendment holding of Warner Communications has been modified by the Court’s more recent Richmond Newspapers decision. Neither the plurality nor concurring opinions in that case suggest that the distinction between access to information and physical access to evidentiary materials is obliterated. Three Justices interpreted that decision as a significant expansion of the First Amendment right of access to information, see 448 U.S. at 583, 100 S.Ct. at 2831 (Stevens, J., concurring); id. at 586, 100 S.Ct. at 2833 (Brennan, J., concurring); id. at 604, 100 S.Ct. at 2842 (Blackmun, J., concurring), but their opinions together received only four votes. The holding of the case is considerably narrower. In his plurality opin ion, the Chief Justice relied on traditional First Amendment doctrine to find only that the Constitution does not permit a criminal trial to be closed to the public by the agreement of the trial judge and the parties, absent some showing of special need. He explicitly noted that the Court did not decide whether the First Amendment afforded access rights to civil trials, id. at 580 n.17, 100 S.Ct. at 2829 n.17, or under what circumstances a criminal trial could be closed, id. at 581 n.18,100 S.Ct. at 2830 n.18.

In sum, we do not think that Richmond Newspapers impliedly diminishes the explicit holding of Warner Communications that there is no constitutional right to inspect and copy. 101

D. First Amendment Interests in the Remaining Documents at Issue

We must now apply these constitutional principles to determine whether the First Amendment requires us to vacate PTO 35 or to unseal documents referred to in the summary judgment hearing, documents referred to in the FPS (250,000 in number including the DSS’s), and documents that contain raw economic data. None of these documents is currently part of the public record; and none of these documents was obtained by plaintiffs other than by use of the discovery process administered by this court. We emphasize that we are not, therefore, confronted with a situation in which a decision not to release these documents would constitute a classic prior restraint. Cf. Nebraska Press Association v. Stuart, supra; Rodgers v. United States Steel Corp., 536 F.2d 1001 (3d Cir. 1976). Because these documents were produced pursuant to PTO 35, the only question before us is whether PTO 35 is a constitutionally valid protective order under the balancing test discussed in Part IV-B, supra. We conclude that it is.

Our first consideration is whether PTO 35 reflects a proper balancing of the need for confidentiality against the First Amendment interest in disseminating discovery materials. We have already enumerated some of the harms that this protective order was intended to prevent, which can be summarized as the danger that discovery would be simply unmanageable in this litigation. We stress that in a case of this size and complexity, that danger is substantial and unmanageable discovery creates serious delays and a risk of unfairness. An additional potential harm to which PTO 35 responds is the risk of damaging disclosure of commercial information about defendants. In view of the liberal discovery allowed by the federal rules, we think that this risk is also serious. We conclude that the likelihood of harm under the circumstances of this case clearly outweighs plaintiffs’ interest in communicating information gleaned only from discovery materials.

Next we must ask whether a more narrowly drawn order or a less restrictive means than a protective order would effectively meet these concerns. We think it clear that they would not. First, no means other than sealing the documents produced in discovery could protect defendants’ confidential commercial information. 102 Second, PTO 35 is reasonably narrowly drawn. Although it is an umbrella order that permits the parties to designate documents as confidential, nothing other than a blanket order can meet the manageability concerns present in a complex case. Cf. In re San Juan Paper Co., supra, Slip Op. at 117; In re Halkin, supra, 598 F.2d at 196 n.47. The order incorporates an appeals process for challenges to such designations, and by the terms of the order the burden of proving the need for confidentiality remains, as we think the Constitution requires, on the party seeking protection. In addition, the only persons restrained by the order are qualified persons, executives, and witnesses, as defined in PTO 35. We have already noted that the only communication restrained is that of information contained in documents produced under seal.

Finally, the foregoing discussion makes clear why PTO 35 is an effective means to prevent unwarranted exposure of defendants’ commercial information and abuse of the court’s processes, and to insure that discovery would be orderly and fair. We hold that the First Amendment does not require vacatur of PTO 35, nor does it require disclosure of more categories of documents than do the public’s access rights.

V. Conclusion

Our analysis of the facts and the applicable law has led us to the conclusion that we must deny plaintiffs’ motions for wholesale declassification and for vacatur of PTO 35; that we must declassify the DSS’s and the text of the FPS, along with several of its appendices; and that we must refuse to declassify all other materials that were designated “confidential” under the aegis of PTO 35 and either filed of record or accumulated in discovery.

Plaintiffs’ counsel have informed us that a twenty-five volume appendix to plaintiffs’ appellate brief is being filed with the Clerk of the Third Circuit Court of Appeals. We believe that the vast bulk of plaintiffs’ appendix filings consist of materials that are declassified; by virtue of the order accompanying this opinion. If that judgment is correct, many thousands of documents which have never been .on the public record will be exposed at the moment that the Clerk of the Court of Appeals completes the docketing process. Defendants and third parties who have submitted documents in discovery in reliance on PTO 35 and whose documents will be unsealed by virtue of the accompanying order should have an opportunity to make application to the Court of Appeals for a stay and for relief from our decision. 103

Accordingly, we will direct the Clerk promptly to notify third parties that produced documents that are declassified by the order accompanying this opinion of the substance of this decision. 104 We will stay the effective date of our accompanying order until December 15, 1981.

1 . There are eleven other reported opinions in this case. The first three were filed by our immediate predecessor in the case, Judge A. Leon Higginbotham, Jr.: Zenith Radio Corp. v. Matsushita Elec. Indus. Co., 402 F.Supp. 251 (E.D.Pa.), petition denied mem. sub nom. In re Japanese Elec. Prods. Antitrust Litigation, 521 F.2d 1399 (3d Cir. 1975) (constitutionality of antidumping act); Zenith Radio Corp. v. Matsushita Elec. Indus. Co., 402 F.Supp. 262 (E.D. Pa. 1975) (personal jurisdiction and venue); Zenith Radio Corp. v. Matsushita Elec. Indus. Co., 402 F.Supp. 244 (E.D.Pa.), petition denied mem. sub nom. In re Japanese Elec. Prods. Antitrust Litigation, 521 F.2d 1399 (3d Cir. 1975) (international application of Robinson-Patman Act). Our first opinion in the case dealt with the question whether the case was too complex for trial by jury. See Zenith Radio Corp. v. Matsushita Elec. Indus. Co., 478 F.Supp. 889 (E.D.Pa. 1979), vacated, 631 F.2d 1069 (3d Cir. 1980). In addition to the Final Summary Judgment Opinion and the 1916 Antidumping Act opinion referenced in the text, we have filed seven other reported opinions which deal with defendants’ summary judgment motions. We have: (1) denied a motion seeking summary judgment for lack of subject matter jurisdiction, see Zenith Radio Corp. v. Matsushita Elec. Indus. Co., 494 F.Supp. 1161 (E.D.Pa. 1980); (2) denied a motion for summary judgment brought by certain defendants against Zenith on the ground that Zenith was not directly injured and was therefore barred from recovery under the doctrine of Illinois Brick Co. v. Illinois, 431 U.S. 720, 97 S.Ct. 2061, 52 L.Ed.2d 707 (1977), see Zenith Radio Corp. v. Matsushita Elec. Indus. Co., 494 F.Supp. 1246 (E.D.Pa. 1980); (3) denied a motion asserting that National Union Electric Corporation lacked standing to sue under the doctrine of Bangor Punta Operations, Inc. v. Bangor & A. R.R., 417 U.S. 703, 94 S.Ct. 2578, 41 L.Ed.2d 418 (1974), see National Union Elec. Corp. v. Matsushita Elec. Indus. Co., 498 F.Supp. 991 (E.D.Pa. 1980); (4) denied a motion by defendant Sears, Roebuck & Co. for summary judgment based on statute of limitations grounds, see Zenith Radio Corp. v. Matsushita Elec. Indus. Co., M.D.L. No. 189 (E.D.Pa. May 23, 1980) (Pretrial Order 263); and (5) denied a motion by Sharp Electronics Corporation alleging that Zenith’s claims under the 1916 Antidumping Act are barred by the Treaty of Friendship, Commerce and Navigation between the United States and Japan, 4 U.S.T. 2063, T.l.A.S. No. 2863 (1953), see Zenith Radio Corp. v. Matsushita Elec. Indus. Co., 494 F.Supp. 1263 (E.D.Pa. 1980).

Additionally, we filed three opinions dealing with questions of admissibility of documents offered in the pretrial evidentiary hearings. See Zenith Radio Corp. v. Matsushita Elec. Indus. Co., 505 F.Supp. 1313 (E.D.Pa. 1980); Zenith Radio Corp. v. Matsushita Elec. Indus. Co., 505 F.Supp. 1190 (E.D.Pa. 1980); Zenith Radio Corp. v. Matsushita Elec. Indus. Co., 505 F.Supp. 1125 (E.D.Pa. 1980).

2 . See infra pages 874-875.

3 . Our' survey reveals that the number of filings, by category, that have been designated “confidential” by the parties and filed under seal by the clerk is:

Answers to Interrogatories - 174

Deposition Transcripts - 62

Memoranda - 61

Motions - 23

Briefs - 19

Exhibits - 9 •

Objections - 4

Status Reports - 2

Requests for Production - 2

Transcripts of Hearings - 10

Interrogatories - 1

Pretrial Orders - 3

Price Comparisons - 3

Affidavits - 3

Final Pretrial Statements - 9

TOTAL 385

4 . Filing of the defendants’ FPS was deferred pending resolution of the summary judgment motions.

5 . The FPS was required by PTO 154, reprinted at 478 F.Supp. 889, 946 (E.D.Pa.1979), our basic case management order. With respect to the FPS, PTO 154 explains:

By this formulation we do not intend that the parties must provide a script for trial. We do, however, intend that the parties set forth in narrative form not just the ultimate facts, but all the facts they will prove, including all subsidiary or supporting facts. In so doing they will make, at a minimum, the kind of submission they would make if they were writing detailed requests for findings of fact setting forth what the evidence has established in a nonjury case.... In view of the primacy of the conspiracy claims in plaintiffs’ case and defendants’ counterclaims, the FPS shall itemize all overt acts to be proved at trial. In particular, the FPS shall enumerate with specificity the facts (i.e., the evidence aliunde) upon which they rely to prove that each defendant . . . knowingly joined in the alleged conspiracy and all facts, separately as to each defendant, upon which they rely as to each defendant’s participation in the alleged conspiracy. Where any facts will be offered against fewer than all parties, the FPS shall identify the parties against which the facts are or are not offered.

6 . In the July 20, 1977, Memorandum of the Hitachi Defendants Regarding Plaintiffs’ Motion for Wholesale Declassification of Documents, it is estimated that approximately thirty-five million documents were made available to plaintiffs’ counsel by the Japanese defendants, and it is represented that some -77,000 were designated as confidential by the Hitachi defendants alone. Zenith produced millions more for inspection by defendants. While an approximation is difficult, it is certain that at least several million document pages were copied.

7 . See Final Summary Judgment Opinion, 513 F.Supp. at 1119, for a condensed history of the NUE and Zenith litigations and their eventual joinder.

8 . The text of Rule 26(c)(7) is set out in note 39, infra.

9 . “Qualified persons” refers to counsel, their employees, and court personnel.

10 . The parties engaged in extensive negotiations, and submitted proposed drafts to Judge Higginbotham. According to the submission of defendants, see Letter from Kevin T. Hughes to the Court (Jan. 16, 1981), plaintiffs’ conceded the need for a confidentiality order and their proposal differed from defendants’ proposal only with respect to: (1) the procedures to be followed with respect to exposure of documents to designated persons; and (2) which party would bear the burden of proof with respect to disputed confidentiality designations.

11 . For example, Kevin T. Hughes, counsel for Matsushita, stated on the basis of his personal knowledge that the last 100,000 documents produced in discovery by Zenith were produced on paper preprinted with a legend “confidential per court order.” PTO 292, at 247 (Sept. 3, 1980) (Transcript of Hearing on Declassification Motion). These documents obviously had not been pre-reviewed. Plaintiffs have not disputed these assertions.

12 . The motion as originally filed sought an

Order striking defendants’ sweeping designations of confidentiality of all documents, pleadings and materials which were prepared or refer or relate to the period prior to September 21, 1974, the date of the filing of the Complaint by Zenith Radio Corporation.

Plaintiffs’ later submissions appear to abandon the September 21, 1974, line of demarcation. The motion to vacate PTO 35 was filed in May, 1979.

13 . Judge Higginbotham did suggest a method for dealing with the problem. See infra pages 877-878.

14 . Our published opinions in this case, see supra note 1, chronicle the progress of the litigation.

15 . The designation “qualified executive” resulted from an agreement executed by the parties on May 9, 1978, allowing certain executives of the parties to have access to certain documents and information theretofore produced in the litigation for purposes of trial preparation. The agreement listed twelve Zenith qualified executives who were granted access to all documents and information theretofore submitted by defendants in discovery except for certain prescribed price-cost-profit and technical data. The principal limitations related to: (1) access to documents which described defendants’ suggested or published retail prices after September 20, 1974; (2) internal financial statements and tax returns; (3) defendants’ United States Treasury Department submissions; and (4) certain cost and profit data. The agreement provided that the confidential information could be used only in connection with trial preparation; that it was without prejudice to the positions of any party on the confidentiality issue in general; and that the objections under the agreement were reciprocal. The agreement also permitted the Zenith executives to examine Zenith’s preliminary pretrial memorandum, the-precursor of the FPS and our initial effort to secure definition of the issues. Since this agreement has not heretofore been filed of record, we do so now and certify it to the Court of Appeals as a supplement to the record.

The designation “qualified witness” was established by PTO 184 (June 7, 1979) which followed our confidentiality hearing of May 9, 1979 (PTO 172 (May 16, 1979)). PTO 184 made confidential data available to potential trial witnesses and their staff personnel. PTO 184 was the principal limitation upon the access of qualified witnesses to discovery material relating to cost, price, and profit data subsequent to September 20, 1974. Like PTO 35, PTO 184 contained certain “housekeeping” provisions and some limitations on the use of the data in trial preparation. It also contained a specific proscription against disclosure of material to any government agency, lobbyist, the press, or any person who was neither a qualified person, qualified executive, or qualified witness.

16 . We have not acted upon defendants’ counterclaims. Rather, we have certified our grant of summary judgment for appellate review pursuant to Fed.R.Civ.P. 54(b). Zenith Radio Corp. v. Matsushita Elec. Indus. Co., 513 F.Supp. 1334 (E.D.Pa.1981). PTO 311, entered April 8, 1981, by preserving the Court’s jurisdiction to consider a number of matters, has preserved our ability to decide the confidentiality issues, and the parties have agreed that we should do so. See Transcript of Conference Call, Document No. 3482 (April 13, 1981). While we are cognizant of the fact that the record is now before the Third Circuit Court of Appeals (though because of its voluminous nature it remains physically in this court), we believe that adjudication of the confidentiality issues by us at this time is appropriate, in view of: (1) this reservation; (2) the fact that we have had this matter sub judice for quite some time; (3) the fact that the counterclaims to which PTO 35 applies remain before us; and (4) the possibility of remand.

We are unaware of any rules in the Third Circuit governing applications to seal the record. Our understanding of the practice followed in that court over the past several years, applied mostly in the case of grand jury matters, is that matters sealed (or unsealed) in the district court remain in that condition absent application to the court of appeals.

17 . Judge Higginbotham was in the process of assuming a seat on the Third Circuit Court of Appeals.

18 . Having stated that the “confidential” stamp has been used in this case with the greatest liberality, and that counsel (or their paralegals) have designated numerous documents as confidential without any review, we also note that counsel concede that many documents filed un der seal do not merit this treatment. See PTO 292 at 248, 258-59 (Sept. 3, 1980) (Transcript of Confidentiality Argument), and the submissions contained in defendants’ Proposal for Treatment of Their Pre-September 21, 1974 Documents Currently Designated “Confidential” under Pretrial Order No. 35, (July 19, 1977). It is appropriate, however, to add at this juncture that this state of affairs is not without redeeming aspects. The smooth exchange of discovery material is enormously important in the complex case. Wholesale designations without repeated objection facilitate the discovery process. If a judge, magistrate, or special master had to rule upon countless invocations of Rule 26(c)(7), the progress of complex cases would be severely impeded. Indeed, the judge is a veritable hostage; failure to approve an umbrella protection order is a consignment to years of adjudication of the confidentiality of individual documents. Moreover, it is enormously expensive to have thousands of documents reviewed by paralegals and lawyers and to brief and argue their confidentiality to the court. Needless to say, the skyrocketing cost of litigation is a matter of genuine concern to everyone involved in the legal process.

19 . We had on one earlier occasion excluded a spectator from a pretrial hearing. See PTO 222, at 14-15 (Feb. 8, 1980) (Transcript of Pretrial Conference).

20 . As is the usual case with trial exhibits, the DSS’s were submitted to our deputy clerk, who held them in our chambers. They were never formally filed with the Clerk of this Court. The DSS's were for the most part ruled inadmissible, but that fact does not, in our view, affect the outcome. See Part III-D, infra.

21 . Mr. Miyamoto’s telex discusses three categories of documents and the competitive harm to Hitachi which would result from their disclosure. The first category is pricing information. This includes not only the prices of Hitachi products at different distribution levels, but also information pertaining to the schedule of rebates, discounts, promotional allowances, and advertising allowances. Mr. Miyamoto states that, using information of this sort, even from ten years ago, a competitor could estimate the relationship between current prices at Hitachi’s various distribution levels and could contact Hitachi’s distributors and retailers and wean them from Hitachi’s sales network.

The second category concerns factory and distribution costs. By analyzing cost patterns over the years, says Mr. Miyamoto, a competí tor could calculate current costs and discover Hitachi’s strengths and weaknesses.

The third category involves internal memoranda and other data involving licensee, customer, and supplier relationships. Mr. Miyamoto says that Hitachi’s relationships with these people would be damaged if they discovered that they had not received the same treatment which Hitachi had offered to others. Furthermore, he says that disclosure of this information, even if ten or twelve years old, would cause Hitachi this harm because supplier and customer personnel do not change frequently.

22 . Mr. Yamagishi, at the time of his affidavit, was Assistant General Manager of Matsushita’s Legal Affairs Department, Patent and Legal Division. He had also had a good deal of direct experience in Matsushita’s sales and financial divisions where he was involved in the pricing of television, radio and stereo products, as well as the production, sale and marketing of television products. Mr. Yamagishi’s affidavit attempts to explain how pricing data and other categories of internal business information can be used to the competitive disadvantage of an industry member, even if the information antedates 1974. Mr. Yamagishi makes many statements paralleling those made in the Hitachi telex described in note 20, supra. Mr. Yamagishi, however, goes into more detail in explaining how old price and cost data could be used to gain competitive advantage. Mr. Yamagishi also discusses how data covering a long period of time could be analyzed to determine Matsushita’s strategies and the way it reacts to certain stimuli. In his view, disclosure of the information would reveal to a competitor not only the cards in Matsushita’s hand, but how those cards are likely to be played as well. One particular area that concerned Mr. Yamagishi was product development and styling.

23 . Judge Higginbotham denied the United States International Trade Commission’s (ITC) petition for leave to inspect or copy documents subject to PTO 35 and for a writ of mandamus to command Zenith’s and NUE’s compliance with its demand for access to and the right to inspect and copy documents. Relying on GAF Corp. v. Eastman Kodak Co., 415 F.Supp. 129 (S.D.N.Y.1976), Judge Higginbotham reasoned that the ITC should not be permitted to “piggyback” on the discovery produced for use in the instant litigation. He also denied Zenith’s and NUE’s motion for clarification or modification of PTO 35 to permit them to comply with process served on them by the ITC. See PTO 67 (Aug. 18, 1976).

24 . In Zenith Radio Corp. v. Matsushita Elec. Indus. Co., 1978-1 Trade Cas. ¶ 62,019 (E.D.Pa. 1978) (PTO 101), Judge Higginbotham granted a motion by GTE Sylvania, Inc. to quash a subpoena duces tecum served by Zenith and NUE, calling for confidential documents in GTE’s files, which were produced by Zenith and NUE and others to GTE or to the ITC staff, or were prepared by GTE’s counsel in connection with an ITC investigation. Judge Higginbotham did not quash the subpoena as it pertained to non-confidential material in GTE’s possession. He quashed the subpoena as it pertained to documents of Henry Marketing Company, Inc. and Henry Distributing Company, Inc., possessed by GTE. He further denied a Rule 37 motion by Zenith and NUE to compel discovery of those documents relating to the ITC hearing.

Judge Higginbotham reasoned that it would be an abuse of the discovery process to order a defendant in the instant litigation to produce all documents which he had submitted in another case under the judicial imprimatur that those documents, when submitted, were protected as confidential. He also felt that equity considerations should prevent disclosure of even non-confidential documents produced in the ITC investigation. PTO 101 was filed after we had been assigned the case and we concurred in the opinion and order.

25 . In PTO 89 (Nov. 18, 1977), we approved three consent orders executed by Zenith, NUE, the Japanese manufacturing defendants, and the Justice Department, to facilitate a Justice Department Investigation of the plaintiffs’ charges against the defendants. See Final Summary Judgment Opinion, 513 F.Supp. at 1332, n.408.

26 . Informed by Carlson Cos. v. Sperry & Hutchinson Co., 374 F.Supp. 1080 (D.Minn. 1973), we issued a protective order respecting discovery of these documents on the ground that the utility of this discovery would be outweighed by its burden on the parties and on the court. See PTO 163, at 61-63 (April 23, 1979). By virtue of this ruling, it was unnecessary to rule upon the speech or debate claims of United States Senator Robert Morgan of North Carolina who had allegedly received information from Nash and who had intervened to assert his speech and debate claims.

27 . Pursuant to our request, plaintiffs’ counsel has furnished us with a copy of their appendix in the Court of Appeals, and defense counsel have written us explaining their approach to the appendix. It does not appear from these communications that the appendix will contain any substantive evidence plaintiffs did not include in their FPS. •

28 . A spin-off of this point is the motion filed by Dr. Horace DePodwin, plaintiffs’ leading expert, to intervene and to declassify his expert report so that, in other cases, he might use it to answer our criticisms of his report. See Final Summary Judgment Opinion, 505 F.Supp. at 1334-1364. In a separate order, we deny Dr. DePodwin’s motion on the grounds that (1) if his motion is for intervention of right, then it fails because his interests are adequately represented, and (2) if his motion is viewed as one for permissive intervention, then it fails because (a) he does not have an interest in the “main action” of this suit and (b) in our discretion, we deny it because, his interests already adequately represented, Dr. DePodwin’s Motion would only unduly delay adjudication of the rights of the original parties.

We understand Dr. DePodwin’s motivation for seeking to intervene. We are aware that our comments have been used to attack his professional qualifications by counsel in other cases. We have, in the most trenchant terms, dispelled any notion that our ruling on the admissibility of Dr. DePodwin’s report casts any shadow on his personal and professional integrity. On the day on which we granted summary judgment, we condemned the misuse of our evidentiary rulings to attack Dr. DePodwin’s testimony elsewhere. We said at that time, and reaffirm here, that “any law firm which hereafter files such an attack acts in the worst traditions of the bar, acts in dilatory fashion, escalates the cost of litigation, .... [Wjhat I said [about Dr. DePodwin’s report] applied only to this case, and it is my view that these kinds of attacks are frivolous, and I think ought not to take place any more.” See PTO 309, at 6-8 (April 2, 1981) (Transcript of Hearing of March 27, 1981).

29 . Plaintiffs have also asserted that by reason of its breadth, PTO 35 inhibited their trial preparation. We have considered this point, but perceive no inhibition upon plaintiffs’ preparation of their voluminous FPS, especially in the wake of PTO 184 (June 7, 1979). Plaintiffs have at no time pointed to any specific example of inability to develop their case by reason of PTO 35 as amended.

30 . See supra notes 20 and 21.

31 . Plaintiffs vigorously deny the charges of impropriety. While maintaining that they possess the right to publish the documents obtained in discovery, plaintiffs have, on their own initiative, stated that they have no present intention to do so, except in response to appropriate outside inquiries. We reproduce in this regard a portion of a letter to the court from plaintiffs’ lead counsel, Edwin P. Rome:

In response to the Court’s inquiry of last week regarding the use plaintiffs would make of defendants’ documents should Your Honor declassify them, I would respectfully refer the Court to the colloquy on August 29, 1980. The Court inquired what action plaintiffs would take in the event the Court granted the motion to declassify, specifically asking whether Zenith’s public relations staff would deliver the documents to the offices of the Washington Post or to Congressmen from Illinois. I replied that, in my view, although this might well be permissible, I could not say that it would be done:

“I repeat, however, there is no program or intention that I am aware of or have ever heard of by the remotest indirection on the part of Zenith or any of its people to do anything along that line, but I must say to you, I think it would be entirely permissible and even within. First Amendment rights, particularly in connection with the increasing recognition within the Congress of the delicate areas that are involved here.” (PTO 292, at 245-246).

Plaintiffs and plaintiffs’ counsel have no present intention to make such use of any of defendants’ documents, which have been improperly designated ‘confidential’ or are no longer entitled to “confidential” status. We strongly believe, however, that the continuing prior restraint sealing these documents and preventing any public revelation of their contents raises serious First Amendment questions, violates the Federal Rules of Civil Procedure and is unwarranted.

Letter from Edwin P. Rome to the Court (Jan. 2, 1981).

32 . Defendant Sears has asserted that Zenith has breached PTO 35 by making protected material available to the United States Department of Justice and a federal grand jury in Los Angeles, California, in connection with proceedings which led to Sears’ indictment on charges of customs fraud in relation to the importation of television receivers from Japan. The indictment was dismissed, and an appeal is pending before the Ninth Circuit Court of Appeals.

33 . Federal Rule 26(c)(7), and PTO 35, deal only with pretrial discovery, but related legal and equitable doctrines afford protection to confidential information in later proceedings.

34 . For reasons that will be discussed at length, we do not think that plaintiffs’ arguments for wholesale declassification based on the First Amendment and the public’s right of access are persuasive. Any First Amendment interests that plaintiffs may have in information and documents produced in discovery pursuant to a protective order and not filed of record are outweighed by interests favoring continued confidentiality. Similarly, the public’s common law right of access does not attach to materials not made part of the judicial record. These points are discussed in later sections of the opinion, see Parts III and IV infra, but without reference to the subject of wholesale declassification as such.

35 . We note that at the final hearing on the declassification issue counsel for some of the defendants virtually conceded this point. See PTO 292, at 262 (Sept. 3, 1980).

36 . As a preliminary matter, we will discuss whether the pretrial evidentiary hearing was open and identify the concomitant rights, if any, of public access to the documents that were its subject matter.

37 . The FPS and the expert reports make use of the raw economic data, but in these documents the data is digested rather than set forth in an orderly array.

38 . Constitutional questions should not be reached, of course, if a case can be decided on nonconstitutional grounds. Hagans v. Lavine, 415 U.S. 528, 547, 94 S.Ct. 1372, 1384, 39 L.Ed.2d 577 (1974). As will appear infra, we must reach the First Amendment question because of our disposition of the Rule 26(c) and common law access right questions.

39 . Rule 26(c)(7) provides in relevant part:

Upon motion by a party or by the person from whom discovery is sought, and for good cause shown, the court in which the action is pending . . . may make any order which justice requires to protect a party or person from annoyance, embarrassment, oppression, or undue burden or expense, including one or more of the following: ... (7) that a trade secret or other confidential research, development, or commercial information not be disclosed or be disclosed only in a designated way. . . .

40 . In practical terms, it may well be that courts apply a less rigorous standard to consent orders, although we doubt that any judge would approve a consent order not demonstrably rooted in Rule 26(c), both as a matter of judicial integrity and out of concern for potential public access rights, see Part III, infra:

41 . Prior to 1970, the Federal Rules contained no provision for the protection of trade secrets or other confidential commercial information subject to discovery in litigation. The predecessor of Rule 26, Rule 30(b), afforded protection from disclosure only to “secret processes, developments or research.” Despite the failure of the Rules to include trade secrets and confidential information among the types of information which could be protected, federal courts found ways of protecting such information. Frequently, this was accomplished by analogizing trade secrets and confidential commercial information to the types of information explicitly covered by Rule 30(b) or the other subject matter which had traditionally been protected in a court’s discretion. See, e.g., Covey Oil Co. v. Continental Oil Co., 340 F.2d 993 (10th Cir.), cert. denied, 380 U.S. 964, 85 S.Ct. 1110, 14 L.Ed.2d 155 (1965); Essex Wire Corp. v. Eastern Elec. Sales Co., 48 F.R.D. 308 (E.D.Pa. 1969). The addition of Rule 26(c)(7) in 1970 reflected the existing law that courts would limit the disclosure of trade secrets and confidential commercial information revealed in discovery, see Advisory Commission Notes to Rule 26(c)(3), reprinted in 48 F.R.D. 485, 497-508 (1970), and obviated any need of comparison with other types of information by recognizing that trade secrets and confidential commercial information are subject matter protectible in their own right. Because Rule 26(c)(7) reflected existing law under old Rule 30(b), we may look to cases decided under that Rule in considering the instant issues.

42 . See, e.g., Magnavox Co. v. Mattel, Inc., No. 80-4124 (N.D.Ill. March 24, 1981) (agreements between patentee and licensee, patent sub-license agreements, and royalty reports); Chesa Int’l, Ltd. v. Fashion Assocs., Inc., 425 F.Supp. 234 (S.D.N.Y.), aff’d mem., 573 F.2d 1288 (2d Cir. 1977) (customer list); Vollert v. Summa Corp., 389 F.Supp. 1348 (D.Hawaii 1975) (financial records detailing capitalization, net worth, and annual income); Maritime Cinema Serv. Corp. v. Movies En Route, Inc., 60 F.R.D. 587 (S.D.N.Y.1973) (license fees and oral contracts with customers); Spartanics, Ltd. v. Dynetics Eng'r. Corp., 54 F.R.D. 524 (N.D.Ill. 1972) (information pertaining to market entry); Russ Stonier, Inc. v. Droz Wood Co., 52 F.R.D. 232 (E.D.Pa. 1971) (customer and supplier list); Corbett v. Free Press Assoc., 50 F.R.D. 179 (D.Vt.1970) (profit and gross income data); Essex Wire Corp. v. Eastern Elec. Sales Co., supra, (terms of contract); Hecht v. Pro-Football, Inc., 46 F.R.D. 605 (D.D.C.1969) (financial statements); Borden Co. v. Sylk, 289 F.Supp. 847 (E.D.Pa. 1968), appeal dismissed, 410 F.2d 843 (3d Cir. 1969) (prices charged and volume sold to customer); Turmenne v. White Consol. Indus., Inc., 266 F.Supp. 35 (D.Mass.1967) (customer lists); American Oil Co. v. Pennsylvania Petrol. Prods. Co., 23 F.R.D. 680 (D.R.I.1959) (lists of dissatisfied customers).

43 . “Very serious injury” was required in United States v. International Business Mach. Corp., 67 F.R.D. 40 (S.D.N.Y.1975). Accord, Citicorp v. Interbank Card Ass’n, 478 F.Supp. 756 (S.D.Ñ.Y.1979); Reliance Ins. Co., supra. We question the appropriateness of and necessity for this higher standard, under Rule 26(c)(7). We defer our consideration of what the First Amendment may require to Part IV, infra.

The “very serious injury” standard of United States v. IBM may be limited to the particular facts of that case. Judge Edelstein was confronted with a request that disclosure be limited in the face of the Publicity in Taking Evidence Act, 15 U.S.C. § 30 (1970), which mandated that depositions taken in a suit brought by the United States under the antitrust laws be open to the public. That statute is of no effect in the case before us.

44 . Compare Timken Co. v. United States Customs Serv., No. 79-2545 (D.D.C. April 25, 1981) (Freedom of Information Act exemption case): four affidavits established a sufficient basis for the court to find that release of pre-1973 information in 1981 would cause substantial harm to the competitive position of the companies from which it was obtained.

It would allow competition to discern the strengths and weaknesses of the marketing strategies of these companies and target their weak points for attack. Competitors also could imitate the successful policies of these companies. Further, customer relations likely would be disrupted by the breach of confidentiality and increased competition from competitors.

Id., Slip Op. at 2.

45 . See notes 21 and 22 supra.

46 . See cases cited in note 41 supra.

47 . See Federal Rule 45(d) (a person to whom a subpoena is issued by a court in district other than that in which the suit is pending may restrain the deposition by filing an objection with the issuing court).

Judge Higginbotham faced a related problem earlier in this case when the Truesdell Distributing Corporation, a Zenith distributor, was served by the District Court of Nebraska with a subpoena to appear and bring specified documents to a deposition. When Truesdell refused to comply with the subpoena, Judge Higginbotham refused to hold the corporation in contempt for lack of jurisdiction. Only the Nebraska court had the requisite jurisdiction. See PTO 84 (Dec. 1, 1977).

48 . Plaintiffs have cited us to no precedent approving of wholesale declassification.

49 . Judge Smith’s opinion consists of but three paragraphs worth reproducing here:

The purpose of this litigation is to determine whether the defendants are liable to the plaintiffs under the antitrust laws and, if so, in what amount. That in itself is difficult enough. The protective order was issued so that I would be spared the duty of deciding applications for protective orders during the course of the discovery. Massive quantities of documents have been furnished by defendants under the umbrella of the protective order and I have been spared such problems.

In deciding the motion I have three alternatives. I might vacate the protective order in toto. There would be some element of a breach of faith in this since the discovery was had under the protection of it and conceivably there may be in the documents or depositions matters which are for some reason properly the subject of a protective order. I have, therefore, rejected this alternative. As a second alternative I might do what California suggests and require that defendants now justify the protective order as to each item which they believe should be protected. Some person would then be. required to pass upon those justifications. I do not propose to be that person. I do not propose in this litigation to referee discovery disputes between the Federal Trade Commission, a stranger to this litigation, and the defendants. I, therefore, reject this alternative.

The third alternative, which is the one I choose, maintains the status quo. The plaintiffs have had a broad discovery and the facts revealed by it are available to them. That was the purpose of the discovery to begin with. If the Federal Trade Commission has a right to the information revealed by discovery in this case, it may pursue that right in whatever forum is proper.

50 . Plaintiffs in this case unquestionably have standing to assert the access rights of the public, because they are themselves members of the public.

51 . See Richmond Newspapers, Inc. v. Virginia, 448 U.S. 555, 564, 569, 100 S.Ct. 2814, 2821, 2823, 65 L.Ed.2d 973 (1980) (plurality opinion) (tradition of public trials inheres in this country’s organic laws); id. at 591, 100 S.Ct. at 2835 (Brennan, J., concurring) (Supreme Court “has persistently defended the public character of the trial process”).

52 . Unlike the English courts, American courts have not linked the right of access to showings of special needs or interests. Justice Powell explained in Nixon v. Warner Communications, Inc. that

The interest necessary to support the issuance of a writ compelling access has been found, for example, in the citizen’s desire to keep a watchful eye on the workings of public agencies, and in a newspaper publisher’s intention to publish information concerning the operation of government.

435 U.S. at 597, 98 S.Ct. at 1311-12 (citations omitted).

53 . Justice Brennan, in his concurring opinion joined by Justice Marshall, similarly reviewed the practice at common law of open trials, agreeing that public access furthered just adjudication and served “broadly political” purposes: to “demonstrate [ ] the fairness of the law to our citizens,” 448 U.S. at 594, 100 S.Ct. at 2837; to act as a check on one branch of government, “akin in purpose to the other checks and balances that infuse our system of government,” id. at 596, 100 S.Ct. at 2838; and to aid accurate factfinding, in the interests of the public as well as the litigants, id.

54 . The Cianfrani court considered the constitutionality of closing pretrial suppression hearings to the public, and concluded that the Constitution included “a strong presumption of public access.” 573 F.2d at 851. Though the majority opinion relied on the Sixth Amendment, and the Supreme Court subsequently decided that the Sixth Amendment did not incorporate such public access rights, see Gannett Co. v. DePasquale, 443 U.S. 368, 99 S.Ct. 2898, 61 L.Ed.2d 608 (1979), Chief Judge Seitz’ discussion of the Anglo-American tradition of judicial openness certainly remains vital. In United States v. Criden, supra, 648 F.2d at 819 n.5, the Third Circuit noted that Judge Gibbons’ concurring opinion in Cianfrani relied on the First Amendment and that the Gannett court had not decided whether that amendment afforded the public access to pretrial hearings.

55 . The case before us did not proceed to trial, since we granted summary judgment for defendants and reserved any decision on defendants’ counterclaims. Although all proceedings were technically pretrial proceedings, we believe that pretrial evidentiary hearings in which rulings binding at trial are made, can fairly be said to be part of the trial itself. In any event, we have no doubt that the presumption of openness applies as fully to pretrial as to trial proceedings, at least in civil litigation.

With respect to access rights to criminal proceedings, the Supreme Court has distinguished between trial and pretrial proceedings. In Gannett Co. v. DePasquale, the court held that the press and public could constitutionally be excluded from a pretrial suppression hearing in a murder prosecution. The Court concluded that the Sixth Amendment did not incorporate the common law rule affording access rights to the public, but that even if it could be understood to do so, “there exists no persuasive evidence that at common law members of the public had any right to attend pretrial proceedings; indeed, there is substantial evidence to the contrary.” 443 U.S. at 387, 99 S.Ct. at 2909 (footnote omitted). But in Cianfrani, the suggestion was made that the public’s right of access to pretrial suppression hearings is found in the “federal common law implied from the first amendment.” 573 F.2d at 862 (Gibbons, J., concurring). The year following the Gannett decision, the Court decided that the First Amendment did incorporate the common law right of the public to access to criminal trials. Richmond Newspapers, Inc. v. Virginia, supra. We doubt that the Court’s differentiation between trial and pretrial proceedings in the criminal context is instructive in the civil context because Gannett Co. v. DePasquale construed only the Sixth Amendment. Moreover, the access cases in the criminal context are characterized by an overriding concern about prejudicing a prospective jury panel. This concern is rarely present in civil cases, and certainly not in this one. Any differentiation, moreover, is not relevant to our determination whether access rights have attached to certain materials. The attachment of the public’s right to inspect and copy judicial records depends not on whether the proceeding recorded was “trial” or “pretrial,” but whether it was public. Putting it differently, the question before us is what materials are judicial records in the public domain. Cf. Cox Broadcasting Corp. v. Cohn, 420 U.S. 469, 491-96, 95 S.Ct. 1029, 1044 — 47, 43 L.Ed.2d 328 (1975) (in deciding First Amendment validity of imposition of civil liability for publication of rape victim’s name, fact that information was taken from court records in public domain was critical).

56 . A 1980 amendment to Rule 5(d) provides that “the court may on motion of a party or on its own initiative order that depositions upon oral examination and interrogatories, requests for documents, requests for admission, and answers and responses thereto not be filed unless on order of the court or for use in the proceeding.” Such discovery materials not filed would then not be accessible by the public. See Advisory Committee Notes to Rule 5(d), reprinted in 85 F.R.D. 521, 524 (1980) (filing materials favored because public may have no access otherwise).

57 . The right to copy, which the Third Circuit has said is “correlative” to the right to inspect, includes the right to copy by mechanical means. United States v. Criden, supra, 648 F.2d at 823.

58 . We understand Judge Wisdom’s phrase “introduced into evidence” to mean offered rather than received into evidence. See infra.

59 . This conclusion is not changed if the fruits of discovery are physically but not officially within the custody of the court. We have described in an earlier opinion how a room in this courthouse was converted into a document depository for the convenience of the parties and the court. See Final Summary Judgment Opinion, supra, 513 F.Supp. at 1132. In no sense were the documents stored therein officially within the custody of the district court; indeed, the lock on the door was changed, and counsel were given keys.

60 . We will discuss possible public access rights under the First Amendment in Part IV, infra.

61 . The Hubbard court also indicated that deciding that certain materials or documents are part of the case record does not end the inquiry whether they are or should be part of the public case record. 650 F.2d at 297 n.ll. The documents at issue in the Hubbard case were introduced under seal in a pretrial suppression hearing. Though part of the case record, they were therefore not part of a public record until the entry of the unsealing order that was the subject of the Hubbard appeal.

62 . Appellate Rule 10 is instructive but not dis-positive because the determination of what is included in the record on appeal — and what is not — depends in part on the different functions of trial and appellate courts in our legal system. Thus, papers that were filed in the district court but never made part of the proceedings before it may be excluded from the record on appeal because the reviewing court cannot consider them. See, e.g., Busse v. United States, 542 F.2d 421, 427-28 (7th Cir. 1976) (depositions filed in district court after filing of notice of appeal from grant of summary judgment could not be considered on appeal); 9 Moore’s Federal Practice ¶ 210.04[I] (2d ed. 1980). The same papers are nonetheless part of the judicial record accessible by the public if publicly filed.

63 . Of course, such materials are public records only so long as they remain in the custody of the court. But cf. 16 C. Wright, A. Miller, E. Cooper & E. Gressman, Federal Practice and Procedure § 3956, at 386 (1977) (items before district court but subsequently withdrawn by counsel are still part of record on appeal and must be restored). We have serious doubts that a member of the public could acquire access under the common law rule to materials already restored to their owner. We note parenthetically that the Manual for Clerks of United States District Courts requires that exhibits be retained by the clerk in cases of historical significance. Administrative Office of United States Courts, Manual for Clerks of United States District Courts §§ 201.1, 201.2(F)(1) (1978), ch. 13, Ex. 2, at 4-5 (1954).

64 . The Hubbard court considered only the propriety of unsealing papers that were neither part of the stipulated record for trial, nor used in the examination of witnesses or in the trial judge’s opinion, explaining:

[Tjhe Church [of Scientology, a third party] and the individual defendants each acknowledged at oral argument in this court that their positions would have been much “weaker” or “different” if all the documents had been introduced at the suppression hearing in haec verba or if witnesses had been extensively examined concerning the documents or if the trial judge had engaged in a document-by-document analysis in his consideration of the suppression motion. As a practical matter, we take this acknowledgment to abandon claims to the propriety of unsealing at least that group of documents which were used individually in the examination of witnesses at the suppression hearing or in the trial judge’s opinion.

650 F.2d at 302 n.21.

65 . The declassification motion before us requires us to decide also whether documents that are referenced in papers filed on the public record are therefore part of the record for access purposes. We have no difficulty concluding that they are not. To decide otherwise would be to make the record in a complex case like this one into a set of innumerable nesting eggs: a decision to open one would always require a decision whether to open the egg the first contains, the second contains, and so on. The right to inspect and copy is not a license in the public to examine whatever sources counsel drew on in writing their papers.

66 . Accord, Crystal Grower’s Corp. v. Dobbins, 616 F.2d 458, 460-61 (10th Cir. 1980). See our discussion of possible countervailing interests infra.

67 . We also think it important to articulate what we take to be an unspoken exception to the principle that materials merely referred to at a hearing become part of the record. Under this exception, materials offered in bad faith, that is, for the sole purpose of making them public, would be entitled to protection. We emphasize that we do not find this exception applicable in this case.

68 . The existence and importance of this judicial duty are clearly conveyed by the Chief Justice’s plurality opinion in Richmond Newspapers, Inc. v. Virginia. See supra pages 896-897. Former Solicitor General Wade McCree noted recently that

When we read a judicial opinion, we may be swayed in some small measure by whether the writer shares our views or prejudices and concludes with the words “AFFIRMED” as we ourselves would. But it is what comes before — how the issues are stated and how they are resolved — that leads us to conclude whether this is a judge that we are glad to have. Thus, we expect the judge, like no other public official to justify his decisions with reason.

McCree, Bureaucratic Justice: An Early Warning, 129 U.Pa.L.Rev. 777, 780 (1981) (footnote omitted).

69 . The Criden court stressed the importance of the trial court’s careful “articulation of the factors considered and the weight accorded to them.” 648 F.2d at 819.

70 . Under the First Amendment, the press enjoys no greater access rights than the public generally. Nixon v. Warner Communications, inc., 435 U.S. 589, 609, 98 S.Ct. 1306, 1318, 55 L.Ed.2d 570 (1978); Belo Broadcasting Corp. v. Clark, 654 F.2d 423, 428 (5th Cir. 1981). This principle is at least equally applicable to access rights under the common law, which accords no special privileges to the press.

71 . The same question presented in Criden was considered by the Second Circuit in In Re National Broadcasting Co., 635 F.2d 945 (2d Cir.) stay denied, 449 U.S. 895, 101 S.Ct. 261, 66 L.Ed.2d 125 (1980), (United States v. Myers), by the District of Columbia Circuit in In Re National Broadcasting Co., 653 F.2d 609 (D.C. Cir. 1981) (United States v. Jenrette); and by the Fifth Circuit in Belo Broadcasting Corp. v. Clark, supra.

72 . The court of appeals found that the dissemination and rebroadcast to the public of the evidence adduced at the bribery trial of several public officials was necessary to vindicate fully the “public forum values” stressed by the Supreme Court in Richmond Newspapers. 648 F.2d at 822. The public interest in the proceeding was greater than in the ordinary criminal trial, but only a small number of persons were able personally to view the trial.

In this connection, we think it important to note that although the common law access right seeks to vindicate a general public interest, the subject matter of some proceedings may enhance that interest. Distilling general principles from the reported access cases is not a simple task, because a review of pertinent interests is necessarily closely tied to the facts. Most access cases concern criminal proceedings, moreover, and are primarily concerned with balancing the fair trial rights of criminal defendants with the access rights of the public. See supra note 55.

73 . Compare the holding of the Third Circuit in United States v. Cianfrani:

[A]ny order of exclusion must extend no farther than the circumstances strictly warrant in order to meet the asserted justification of closure. Thus, only that portion of the public may be excluded for only that portion of the proceeding that the court finds to be strictly and inescapably necessary to protect the interests asserted by the defendant in support of his motion to close a hearing subject to the public trial requirement. 573 F.2d at 854 (citation omitted).

74 . As did the Third Circuit in United States v. Criden, the District of Columbia Circuit remanded the Hubbard case for reconsideration in light of its opinion. The court noted in particular that some interests for and against disclosure could not be weighed without examining the documents at issue, and it discussed only what it referred to as “generalized interests.” See United States v. Hubbard, supra, 650 F.2d at 317.

75 . In general, where materials have been accessible previously, the nature of that access is relevant only to a determination whether wider dissemination should or must be permitted. The First Amendment prior restraint doctrine generally bars limitations on existing access. See Nebraska Press Ass'n v. Stuart, 427 U.S. 539, 96 S.Ct. 2791, 49 L.Ed.2d 683 (1976); Cox Broadcasting Co. v. Cohn, supra. The Hubbard court thought, however, that the fact of access during the pendency of the appeal to the documents that were the subject of the appeal was not a factor to be considered on remand. The allowance of such a consideration would only “compound the errors we perceive in the court’s original orders.” 650 F.2d at 318 n.99.

76 . The parties settled their dispute before the court of appeals announced its decision, so no published opinion discussed the documents at issue. 616 F.2d at 460.

77 . Both parties had argued to the appellate court that the record should remain sealed during the pendency of the appeal. Whether the appellate record was ordered to be unsealed after settlement of the dispute on the motion of a party or by the court on its own motion does not appear from the reported opinion. See id.

78 . See also Belo Broadcasting Corp. v. Clark, supra, (refusing to set aside district court order that denied requests to copy tapes admitted as evidence in federal “Brilab” criminal prosecution).

79 . Cf. United States v. Criden, supra, 648 F.2d at 831 (Weis, J., concurring and dissenting). Judge Weis criticized the majority’s effort to articulate the strength of the presumption of access. He thought that the court would do better “simply [to leave] its existence as one factor to be considered.”

80 . We have also observed, supra, that such a hearing is essentially part of the trial.

81 . Indeed, the plaintiffs argue that under the terms of the Japanese Antimonopoly Law the entire record must be considered to be open. Some of the DSS’s, i.e., those containing United States public records and reports, are already in the public domain.

82 . We note additionally that the DSS’s would have to be declassified under the same rationale that we used to declassify the FPS, i.e., that they are records that are the ultimate subject of a dispositive ruling. The close relationship between the DSS’s and the grant of summary judgment is made clear in our Final Summary Judgment Opinion. Although there are some DSS’s on which we did not expressly rule (those relating to the “rebate scheme”) we assumed their admissibility and considered them in connection with the summary judgment motion, so they too must be declassified.

83 . We refer here only to the body of the FPS and not to its non-economic appendices.

84 . We will also declassify the non-economic appendices to the FPS as follows:

Appendix A Plaintiffs Compilation of Japanese Trade Associations and Other Groups

Appendix B Calendar of Conspiratorial Meetings April 12, 1961 - August 7, 1968

Appendix I Compilation of Defendants’ Answers to Interrogatories Relating to Japanese Trade Association and Other Groups

Additionally, Appendix H, setting out Emerson-NUE T.V. Sales and Production data is declassified. Plaintiffs have never asserted that this material is entitled to confidential status. Appendices E and G, which consist of certain record filings and public records, are necessarily released in accordance with Part III-B of this opinion.

85 . We have not ruled expressly upon the sealed filings other than the FPS and the “raw economic data.” That is because the bulk of such material is subsumed within the FPS, and under the circumstances we find the public interest in access to be attenuated. In this particular context the burden upon the court of evaluating each filing creates an interest favoring nondisclosure. Compare our discussion of the propriety of wholesale declassification in Part II-B, supra.

Except for our discussion in Part III — B of discovery materials, we have not addressed that aspect of the motion before us. We take this opportunity simply to restate what we wrote in that section, i.e., that the public has no common law right to inspect materials that are produced in discovery but are not placed in the custody of the court.

86 . The question was reserved by the court in Rodgers v. United States Steel Corp., 536 F.2d 1001 (3d Cir. 1976), which held that a protective order preventing counsel from disseminating information obtained outside the court’s discovery processes was an invalid prior restraint.

87 . See also Reliance Ins. Co. v. Barron’s, supra; Davis v. Romney, supra.

88 . The documents produced by defendants in Halkin v. Helms, No. 75-1773 (D.D.C.), related to the surveillance of anti-Vietnam War activists by the Central Intelligence Agency. Defendants did not seek a protective order under Rule 26(c) prior to complying with plaintiffs’ requests for discovery. Instead, before making these documents available to plaintiffs, defendants deleted all information that might harm the United States’ foreign relations; reveal the investigative methods and sources of the CIA and other agencies; or invade the privacy of third parties. Defendants moved for a Rule 26(c) protective order after they were notified that plaintiffs intended to make public some of the documents produced. See 598 F.2d at 180-SI.

The Halkin decision prompted new attention to the relationship between Rule 26(c) and the First Amendment. See, e.g., Note, Protective Orders Prohibiting Dissemination of Discovery Information: The First Amendment and Good Cause, 1980 Duke L.J. 766; Note, Rule 26(c) Protective Orders and the First Amendment, 80 Colum.L.Rev. 1645 (1980); Case Comment, 92 Harv.L.Rev. 1550 (1979).

89 . In our view, the proposition that mere participation in litigation is not a waiver of First Amendment rights is clear. We do not think that the statement in Rodgers v. United States Steel Corp., supra, 536 F.2d at 1006, that “parties and counsel, by taking advantage of or a part in the discovery processes, [may] implicitly waive their First Amendment rights freely to disclose or disseminate the information obtained through those processes,” suggests that the Third Circuit would hold otherwise. The Rodgers court emphasized that it did not decide if protective orders covering discovery materials obtained through the court’s processes are “ever subject” to the First Amendment’s prohibitions. Rather, it assumed for the purposes of the appeal before it that such orders are constitutional. Because the order it reviewed also restrained the dissemination of information obtained independently of the court, the order was found to be an unjustifiable prior restraint of speech. Id. at 1008.

Nor do we think that the Second Circuit’s suggestion in International Products Corp. v. Koons, 325 F.2d 403, 407 (2d Cir. 1963) that “we entertain no doubt as to the constitutionality of a rule allowing a federal court to forbid the publicizing, in advance of trial, of information obtained by one party from another by use of the court’s processes,” casts doubt on the proposition that by itself participation in discovery cannot constitute a waiver by parties and counsel of their First Amendment rights. A fair reading of Koons implies at most that those rights may be narrower in the discovery setting than elsewhere.

90 . The order challenged in Halkin was one issued after production of the materials for which protection was sought, but the court apparently would apply this test to protective orders issued prior to discovery. How important the court considered the relative timing of production and protective orders is not clear from the court’s opinion. While noting that plaintiffs have no First Amendment right to information not generally available and charging that defendants confused “plaintiffs’ right of access to materials with restraints imposed on materials after they have been obtained,” the court asserted that discovery could not be conditioned on a waiver of constitutional rights. 598 F.2d at 190.

We think that the timing of the protective order may be important, however, even if the litigants do have significant First Amendment rights in materials produced to them during discovery. First, these rights can be waived if the waiver is knowing and voluntary. See Curtis Publishing Co. v. Butts, supra. Second, in complex cases like this one, the requisites of case management include orderly discovery, as the Halkin court itself realized. 598 F.2d at 196 n.47; see In re San Juan Star Co., supra, at 115. A protective order issued at the beginning of discovery, whether stipulated or granted following a motion by one party, both provides notice and encourages efficient and voluntary production of documents, which contributes to a just outcome. If a protective order, especially one entered with the consent of the parties, is found clearly to constitute a waiver of First Amendment rights, the litigants can assert no right to disclose materials produced after its entry. See National Polymer Prods., Inc. v. Borg-Warner Corp., 641 F.2d 418, 423 (6th Cir. 1981).

91 . The underlying litigation arose from the slaying by Puerto Rican police officers of two political radicals. Chief Judge Coffin characterized this incident as “one of the most controversial and well-publicized events in recent Puerto Rican history,” and explained that the law suit — alleging that the Governor of Puerto Rico and other officials had conspired to arrange the killing — was itself a matter of intense public interest and widespread publicity. The defendants’ depositions, the subject of the protective order, were reported almost at the time given. The order restricted the use only of testimony obtained after its issuance. At 110— 11.

92 . The court did not accept the argument made by Judge Wilkey in Halkin that the power to deny access entirely necessarily includes the lesser power to condition use:

[T]he “lesser included” rationale does not always solve the problem. For example, the police power of a municipality justifies zoning but does not permit preferential zoning for whites only. The power to deny access altogether does not necessarily allow a court to impose any condition upon access that it wishes — particularly when the condition at issue has an impact on separate constitutional concerns.

At 114.

93 . The First Circuit particularly stressed the criteria for admissibility in explaining its conclusion that information acquired through discovery is not speech that is accorded full First Amendment protection. Its emphasis, we think, is rooted in the facts of the case before it. We surmise that defendants’ deposition testimony was likely to contain statements that were highly prejudicial or inflammatory, or that unnecessarily invaded the privacy of nonpanies. The case before us is, of course, very different. Under the circumstances of this complex antitrust litigation, in which we have made and published extensive evidentiary rulings, the threshold for full protection should not be satisfaction of criteria for admissibility under the Federal Rules of Evidence, but use by the parties beyond the discovery stage.

94 . The Halkin court likewise recognized that protective orders promote the successful operation of the civil discovery system: “[a] smoothly operating system of liberal discovery is in the interests of litigants and society as a whole, for it contributes to a full and fair airing of all material facts in controversy.” 598 F.2d at 192 (footnote omitted.)

95 . The court characterized its position as one “midway” between those of the majority and dissenting opinions in Halkin. At 116.

96 . See Note, Rule 26(c) Protective Orders and the First Amendment, supra note 88, at 1653-54, 1657-58 for a discussion of the differences between protective order and prior restraint cases. Within the category of protective order cases, this case is similar to those “symbolic speech” cases which involve the incidental suppression of speech for the purpose of regulating conduct. See Procunier v. Martinez, 416 U.S. 396, 409-14, 94 S.Ct. 1800, 1809-12, 40 L.Ed.2d 224 (1974); United States v. O’Brien, supra.

97 . Compare In re Halkin:

We recognize that flexibility will be required to accommodate the practical needs of the discovery process with the standards enunciated herein, particularly where the discovery embraces a large quantity of documents. It may be appropriate, for example, for a trial court (on a proper showing) to issue a blanket protective order covering all documents in a large-scale exchange of files without prejudice to raising the merits of the protective order as applied to particular documents at a later time.

598 F.2d at 196 n.47.

98 . The incidental restriction on speech must be “no greater than is essential to the furtherance of [the government’s] interest.” United States v. O’Brien, supra, 371 U.S. at 377, 88 S.Ct. at 1679 (1968). The principle that protective orders must be drawn as narrowly as will accomplish their purpose is well established. See, e.g., CBS, Inc. V. Young, 522 F.2d 234, 239-40 (6th Cir. 1975); International Products Corp. v. Koons, supra, 325 F.2d at 408.

99 . Our inability to describe precisely the weight of the First Amendment interest in disseminating discovery materials is somewhat discomfiting. Courts are only beginning to identify the constitutional principles that govern speech rights in such materials. It may be that not only the fact of production through the court’s processes but also the type of speech at issue should determine the weight of the First Amendment interest. The protection afforded expression is dictated in some instances by the type of speech that it represents. See Virginia State Bd. of Pharmacy v. Virginia Citizens Consumer Council, 425 U.S. 748, 96 S.Ct. 1817, 48 L.Ed.2d 346 (1976) (commercial speech). The balance should probably be struck differently, in other words, the showing necessary to support nondisclosure should probably be less, where the dissemination of discovery materials is commercial speech than where it is political speech. And, notwithstanding our conclusion that Judge Wilkey’s dissenting opinion in In re Halkin, supra, does not ascribe sufficient importance to First Amendment rights in discovery materials, it may well be that courts ultimately will decide that the Constitution requires only a modest showing of threatened harm to override them. We share Judge Wilkey’s concern that protective orders should be widely available in civil litigation. These difficult questions need not be decided in this case, however, where the result of balancing interests is clear. See, infra, pages 914-915.

We note also that our balancing test approach does not contemplate calibration of the side of the balance favoring disclosure, in contrast to the side of the balance favoring protection. The First Amendment is the interest favoring disclosure. Though the interest in disclosure may seem relatively weak here, in view of plaintiffs’ ability to disseminate the huge amount of material declassified by this Opinion and Order, such First Amendment rights as exist in given material do not depend on the right or ability of the party asserting them to publish other similar materials.

100 . The press generally enjoys no greater access rights to information than does the public because the press is viewed as a representative of the public for the purpose of First Amendment analysis. See, e.g., Nixon v. Warner Communications, Inc., supra, 435 U.S. at 609-10, 98 S.Ct. at 1317-18; Saxbe v. Washington Post Co., 417 U.S. 843, 849-850, 94 S.Ct. 2811, 2814-15, 41 L.Ed.2d 514 (1974).

101 . The Fifth Circuit Court of Appeals has also decided that Warner Communications’ First Amendment holding is not disturbed by Richmond Newspapers. Belo Broadcasting Corp. v. Clark, supra, 654 F.2d at 426-29. We do not think that footnote 6 of the Criden majority opinion suggests that the Third Circuit would reach a different conclusion. The footnote reads:

Arguably, the Richmond Newspapers case could be viewed as supporting a right of the public to access to the tapes through the medium of the broadcasters. The only First Amendment right rejected in Warner Communications was that possessed by the press. The Court stated that “[t]he First Amendment generally grants the press no right to information about a trial superior to that of the general public,” 435 U.S. at 609, 98 S.Ct. at 1318, but did not discuss the nature of the right, if any, of the general public under the First Amendment. In view of our disposition of this case on nonconstitutional grounds, we need not decide whether there remains any room for a constitutional right based on the First Amendment as construed in Richmond Newspapers.

648 F.2d at 821 n.6. We believe that this passage reflects no more than a careful exegesis of the issues in the Criden appeal. Moreover, Judge Weis, concurring and dissenting in Criden, specifically noted that the right to inspect and copy is not of constitutional stature, and that Richmond Newspapers addressed a different issue. Id. at 830.

102 . This case is different from cases where the harm sought to be avoided is the infringement of a defendant’s right to a fair trial, and there are means other than limiting communication that may insure a fair and impartial jury. See Nebraska Press Association v. Stuart, supra, 427 U.S. at 563-65, 96 S.Ct. at 2804-05. But see In re San Juan Paper Co., supra, at 117-119 (finding entry of protective order justifiable to guard right to fair trial). Where, as here, the harm sought to be prevented is communication, only a protective order will suffice.

103 . It may also be that the plaintiffs will be aggrieved by our failure to declassify certain material and will wish to seek relief.

104 . Counsel for plaintiffs and liaison counsel for defendants have supplied this court with a list of such third parties. This list and an appropriate letter of notification will be filed of record with this opinion.