dissenting:
I respectfully dissent. Chief Judge Winter for the majority has written fairly and comprehensively of the conflicting public interests underlying the issues in this appeal. I disagree, however, with the majority’s resolution of the conflict.
I
In my view, the principal issue is the same as that presented in Palmieri v. State of New York, where the Second Circuit observed that “[t]his case lies at the juncture of two competing imperatives in our judicial system: the need of a ... grand jury to gather evidence for ongoing criminal investigations, and the need of our district courts and civil litigants to facilitate efficient resolution of disputes ...” 779 F.2d 861, 864 (2d Cir.1985). Palmieri upheld a protective order against a government attack because, among other reasons, litigants in our civil justice system must be able to rely on assurances of confidentiality given by trial judges.1 The Second Cir*1479cuit’s longstanding emphasis on the importance to civil discovery of upholding protective orders reflects my views.
[T]he vital function of a protective order issued under Rule 26(c), F.R.Civ.P. ... is ‘to secure the just, speedy and inexpensive determination’ of civil disputes, Rule 1, F.R.Civ.P., by encouraging full disclosure of all evidence that might conceivably be relevant. This objective represents the cornerstone of our administration of civil justice. Unless a valid Rule 26(c) protective order is to be fully and fairly enforceable, witnesses relying upon such orders will be inhibited from giving essential testimony in civil litigation, thus undermining a procedural system that has been successfully developed over the years for disposition of civil differences.
[Ajbsent a showing of improvidence in the grant of a Rule 26(c) protective order or some extraordinary circumstance or compelling need ... a witness should be entitled to rely upon the enforceability of a protective order against any third parties, including the Government, and ... such an order should not be vacated or modified merely to accommodate the Government’s desire to inspect protected testimony for possible use in a criminal investigation....
Martindell v. International Telephone & Telegraph, 594 F.2d 291, 295, 296 (2d Cir.1979); see also Palmieri, 779 F.2d at 864-865 (and cases cited therein).
The deponents here would not have provided the deposition testimony but for their reliance on the district court’s statement that the testimony would be beyond the government’s reach. In balancing the grand jury interest against the interest in supporting the integrity of a judicial guarantee and the purposes promoted by Rule 26,1 am not persuaded that the grand jury interest here establishes a compelling need to breach the protective order.
The majority eschews discrete balancing of competing interests and creates a per se rule overriding Rule 26(c) protective orders when opposed by grand jury subpoenas. It reaches this result by balancing abstract concepts rather than the facts of each case. In the process I feel it has swept away a long-standing tool for the management of civil litigation. Litigants relying on protective orders offer a significant amount of evidence in civil cases, but our ruling today will squelch the desire of future deponents to give evidence under a protective order. No longer will deponents be able to rely on the power of a United States District Court to limit disclosure of their testimony to third parties. The government will become a new player in the civil discovery process, forcing deponents to invoke the fifth amendment and thus retarding effective and informative discovery.
II
While agreeing generally with what the majority has written concerning the invocation of the fifth amendment privilege, I nevertheless feel that the outcome it reaches permits an end run by the government around these deponents’ privileges against self-incrimination. The deponents possessed information that would help resolve civil litigation but refrained from disclosing it because of their apprehension that it would be transmitted to the grand jury. They testified only in reliance on the district court’s pledge to protect their potentially self-incriminating remarks from the grand jury.
The grand jury has awesome investigatory powers and, as the majority correctly notes, it operates on the principal that the public has a right to everyman’s evidence. Branzburg v. Hayes, 408 U.S. 665, 688, 92 S.Ct. 2646, 2660, 33 L.Ed.2d 626 (1972). Rather than counterbalancing the use of protective orders as suggested by the majority, however, the great reservoir of power reposed in the grand jury represents an additional reason for enforcing such orders.2
*1480Here, without calling the deponents to appear, the grand jury will obtain evidence given in express contemplation of protection from that body. The government receives evidence it would not have obtained but for overriding a protective order that spawned the deposition in question. The deponents in agreeing to the protective order waived fifth amendment rights in reliance on the order of a district court, and now face the possibility of criminal action based on evidence the grand jury otherwise could not have obtained without their having been immunized.
In the majority view, the enforcement of a protective order against a grand jury subpoena for evidence would constitute a judicial grant of immunity. I agree that granting immunity from prosecution in exchange for testimony is the exclusive prerogative of the executive branch.3 Protective orders, however, are limited judicial devices to foster quick and productive discovery. Testimony given under Rule 26(c) is isolated and apart from a criminal investigation and does not alter the deponent’s potential culpability. Rule 26 does not usurp powers of the grand jury and prosecutor, but instead reflects the authority of a trial judge to manage discovery and control access to judicial records. See Seattle Times Co. v. Rhinehart, 467 U.S. 20, 36, 104 S.Ct. 2199, 2209, 81 L.Ed.2d 17 (1984).
The majority’s argument that a protective order “may seriously impede a criminal investigation by a grand jury” both overstates the reach of the Rule 26(c) shield and understates the retained power of the grand jury. Rule 26(c) provides only limited protection for evidence that, here, would not exist but for the protective order itself. Unlike a grant of immunity, a protective order has no effect on the continued conduct of the grand jury’s investigation. The government remains free to call the deponents before the grand jury to explore their fifth amendment claims and to prosecute or grant immunity. If the deponents are immunized, the grand jury can then demand the sought-after evidence and the balance would tilt conclusively towards enforcing a subpoena for previous deposition testimony.
Another reason that protective orders fall short of a grant of immunity is that they can be modified. Perjury committed by deponents under Rule 26(c) protection, for example, would obviously justify modifying the protective order to permit enforcement of a subpoena for depositions given under it. The potential for modification of a protective order or overriding it through a subpoena thus belies fears of harming grand jury investigations. Here, Judge Hilton expressly provided that his order was subject to modification.
Ill
The majority argues that Rule 26(c) is ineffective and should be supplanted by *1481other methods of eliciting evidence from recalcitrant witnesses. I disagree. The vitality of the protective order has made it a valuable method of facilitating discovery.
Placing higher burdens of proof upon those who invoke the fifth amendment in civil litigation could well retard the truth-seeking function of civil discovery, and it unnecessarily penalizes the exercise of a constitutional right. Nor are stays of discovery during a grand jury investigation and court inquiries into the validity of fifth amendment claims adequate substitutes for Rule 26(c) protective orders. While these are both useful trial management techniques, they can not replace the longstanding role played by Rule 26(c). In my view, Judge Hilton was well within appropriate discretion in rejecting the deponents’ motion for a stay because of the pressing need for the depositions in question. See D’Ippolito v. American Oil Co., 272 F.Supp. 310, 312 (S.D.N.Y.1967). I also disagree that possible leaks of confidences covered by a protective order and the potential for disclosure of information at trial warrant the per se rule announced by the majority. Those factors form valid tactical concerns for parties considering protective orders during discovery, but the very purpose of Rule 26(c) is to foster wide-ranging, successful discovery that will often forestall a public trial.
In sum, I feel that Rule 26(c) has worked long and quietly to effect the efficient and just management of civil discovery. The emasculation of protective orders by today’s decision replaces the rule with a new presence in civil litigation — the roving eye of government and the attendant rush by affected civil litigants to invoke the fifth amendment in discovery.
. The potentially self-incriminating deposition testimony at issue here makes this a stronger case for denying the government access than presented in Palmieri. In that case, the grand jury issued a subpoena ad testificandum and the government sought a modification of the protective order to require Palmieri to testify concerning a protected settlement agreement and to gain access to- the settlement negotiations and terms.
. As the Seventh Circuit has observed, "the explicit grant of such extensive investigative powers should be construed to preclude the implication of supplemental powers, absent unusual circumstances.” Wilk v. American Medical Association, 635 F.2d 1295, 1300 (7th Cir.1980); *1480see also H.L. Hayden Co. v. Siemens Medical Systems, 106 F.R.D. 551, 556 (S.D.N.Y.1985), aff'd, 797 F.2d 85 (2d Cir.1986).
. The majority is concerned that enforcement of a protective order to preclude a grand jury's access to the protected materials would constitute unseemly "judicial intervention into executive prerogatives.” To the contrary, there is no interference with executive processes when a judge enters a protective order prior to any indication that a grand jury may desire to review the protected materials. Likewise, the universe of information to which the fifth amendment privilege could potentially attach may not be, or may never be, the subject of a grand jury investigation. The majority’s separation of powers logic, moreover, encourages prosecutorial intrusion into the judicial domain by validating the use of a subpoena duces tecum as a per se enforceable discovery device. What prosecutor will now invoke the less intrusive process of seeking modification of a protective order from the issuing judge when he knows he has virtually unreviewable authority to override the judge’s order simply by shielding himself in the cloak of the grand jury’s “broad investigative function?” Not only does the subpoena pit the executive against the judicial, but it may result, as it did in this case, in conflict within the judicial branch. For enforcement, the prosecutor must appear before one district judge and ask him to override the issuing judge's order. Here, Judge Hargrove effectively "reversed” the decision of Judge Hilton. In contrast, seeking modification from the issuing judge and appealing, if necessary, an adverse decision, would conform with common trial and appellate practice, and avoids the friction inherent in the subpoena procedure. See United States v. GAF Corp., 596 F.2d 10, 16 (2d Cir.1979).