524 U.S. 399
SCOTUS
1998

Lead Opinion

Chief Justice Rehnquist

delivered the opinion of the Court.

Petitioner James Hamilton, an attorney, made notes of an initial interview with a client shortly before the client’s death. The Government, represented by the Office of Independent Counsel, now seeks his notes for use in a criminal investigation. We hold that the notes are protected by the attorney-client privilege.

This dispute arises out of an investigation conducted by the Office of the Independent Counsel into whether various individuals made false statements, obstructed justice, or committed other crimes during investigations of the 1998 dismissal of employees from the White House Travel Office. Vincent W. Poster, Jr., was Deputy White House Counsel when the firings occurred. In July 1993, Poster met with petitioner Hamilton, an attorney at petitioner Swid-ler & Berlin, to seek legal representation concerning possible congressional or other investigations of the firings. During a 2-hour meeting, Hamilton took three pages of *402handwritten notes. One of the first entries in the notes is the word “Privileged.” Nine days later, Foster committed suicide.

In December 1995, a federal grand jury, at the request of the Independent Counsel, issued subpoenas to petitioners Hamilton and Swidler & Berlin for, inter alia, Hamilton’s handwritten notes of his meeting with Foster. Petitioners filed a motion to quash, arguing that the notes were protected by the attorney-client privilege and by the work-product privilege. The District Court, after examining the notes in camera, concluded they were protected from disclosure by both doctrines and denied enforcement of the subpoenas.

The Court of Appeals for the District of Columbia Circuit reversed. In re Sealed Case, 124 F. 3d 230 (1997). While recognizing that most courts assume the privilege survives death, the Court of Appeals noted that holdings actually manifesting the posthumous force of the privilege are rare. Instead, most judicial references to the privilege’s posthumous application occur in the context of a well-recognized exception allowing disclosure for disputes among the client’s heirs. Id., at 231-232. It further noted that most commentators support some measure of posthumous curtailment of the privilege. Id., at 232. The Court of Appeals thought that the risk of posthumous revelation, when confined to the criminal context, would have little to no chilling effect on client communication, but that the costs of protecting communications after death were high. It therefore concluded that the privilege was not absolute in such circumstances, and that instead, a balancing test should apply. Id., at 233-234. It thus held that there is a posthumous exception to the privilege for communications whose relative importance to particular criminal litigation is substantial. Id., at 235. While acknowledging that uncertain privileges are disfavored, Jaffee v. Redmond, 518 U. S. 1, 17-18 (1996), the Court of Appeals determined that the uncertainty introduced by its balancing test was insignificant in light of existing excep*403tions to the privilege. 124 F. 3d, at 235. The Court of Appeals also held that the notes were not protected by the work-produet privilege.

The dissenting judge would have affirmed the District Court’s judgment that the attorney-client privilege protected the notes. Id., at 237. He concluded that the common-law rule was that the privilege survived death. He found no persuasive reason to depart from this accepted rule, particularly given the importance of the privilege to full and frank client communication. Id., at 237.

Petitioners sought review in this Court on both the attorney-client privilege and the work-product privilege.1 We granted certiorari, 523 U. S. 1045 (1998), and we now reverse.

The attorney-client privilege is one of the oldest recognized privileges for confidential communications. Upjohn Co. v. United States, 449 U. S. 383, 389 (1981); Hunt v. Blackburn, 128 U. S. 464, 470 (1888). The privilege is intended to encourage “full and frank communication between attorneys and thefr clients and thereby promote broader public interests in the observance of law and the administration of justice.” Upjohn, supra, at 389. The issue presented here is the scope of that privilege; more particularly, the extent to which the privilege survives the death of the client. Our interpretation of the privilege’s scope is guided by “the principles of the common law ... as interpreted by the courts ... in the light of reason and experience.” Fed. Rule Evid. 501; Funk v. United States, 290 U. S. 371 (1933).

The Independent Counsel argues that the attorney-client privilege should not prevent disclosure of confidential communications where the client has died and the information is relevant to a criminal proceeding. There is some authority for this position. One state appellate court, Cohen v. Jenkintown Cab Co., 238 Pa. Super. 456, 357 A. 2d 689 (1976), *404and the Court of Appeals below have held the privilege may be subject to posthumous exceptions in certain circumstances. In Cohen, a civil case, the court recognized that the privilege generally survives death, but concluded that it could make an exception where the interest of justice was compelling and the interest of the client in preserving the confidence was insignificant. Id., at 462-464, 357 A. 2d, at 692-693.

But other than these two decisions, cases addressing the existence of the privilege after death-most involving the testamentary exception — uniformly presume the privilege survives, even if they do not so hold. See, e. g., Mayberry v. Indiana, 670 N. E. 2d 1262 (Ind. 1996); Morris v. Cain, 39 La. Ann. 712, 1 So. 797 (1887); People v. Modzelewski, 611 N. Y. S. 2d 22, 203 A. 2d 594 (App. Div. 1994). Several State Supreme Court decisions expressly hold that the attorney-client privilege extends beyond the death of the client, even in the criminal context. See In re John Doe Grand Jury Investigation, 408 Mass. 480, 481-483, 562 N. E. 2d 69, 70 (1990); State v. Doster, 276 S. C. 647, 650-651, 284 S. E. 2d 218, 219 (1981); State v. Macumber, 112 Ariz. 569, 571, 544 P. 2d 1084, 1086 (1976). In John Doe Grand Jury Investigation, for example, the Massachusetts Supreme Judicial Court concluded that survival of the privilege was “the clear implication” of its early pronouncements that communications subject to the privilege could not be disclosed at any time. 408 Mass., at 483, 562 N. E. 2d, at 70. The court further noted that survival of the privilege was “necessarily implied” by cases allowing waiver of the privilege in testamentary disputes. Ibid.

Such testamentary exception cases consistently presume the privilege survives. See, e. g., United States v. Osborn, 561 F. 2d 1334, 1340 (CA9 1977); DeLoach v. Myers, 215 Ga. 255, 259-260, 109 S. E. 2d 777, 780-781 (1959); Doyle v. Reeves, 112 Conn. 521, 152 A. 882 (1931); Russell v. Jackson, 9 Hare 387, 68 Eng. Rep. 558 (V. C. 1851). They view testa*405mentary disclosure of communications as an exception to the privilege: “[T]he general rule with respect to confidential communications ... is that such communications are privileged during the testator’s lifetime and, also, after the testator’s death unless sought to be disclosed in litigation between the testator’s heirs.” Osborn, 561 F. 2d, at 1340. The rationale for such disclosure is that it furthers the client’s intent. Id., at 1340, n. II.2

Indeed, in Glover v. Patten, 165 U. S. 394, 406-408 (1897), this Court, in recognizing the testamentary exception, expressly assumed that the privilege continues after the individual’s death. The Court explained that testamentary disclosure was permissible because the privilege, which normally protects the client’s interests, could be impliedly waived in order to fulfill the client’s testamentary intent. Id., at 407-408 (quoting Blackburn v. Crawfords, 3 Wall. 175 (1866), and Russell v. Jackson, supra).

The great body of this case law supports, either by holding or considered dicta, the position that the privilege does survive in a case such as the present one. Given the language of Rule 501, at the very least the burden is on the Independ*406ent Counsel to show that “reason and experience” require a departure from this rule.

The Independent Counsel contends that the testamentary-exception supports the posthumous termination of the privilege because in practice most cases have refused to apply the privilege posthumously. He further argues that the exception reflects a policy judgment that the interest in settling estates outweighs any posthumous interest in confidentiality. He then reasons by analogy that in criminal proceedings, the interest in determining whether a crime has been committed should trump client confidentiality, particularly since the financial interests of the estate are not at stake.

But the Independent Counsel’s interpretation simply does not square with the ease law’s implicit acceptance of the privilege’s survival and with the treatment of testamentary disclosure as an “exception” or an implied “waiver.” And the premise of his analogy is incorrect, since eases consistently recognize' that the rationale for the testamentary exception is that it furthers the client’s intent, see, e. g., Glover, supra. There is no reason to suppose as a general matter that grand jury testimony about confidential communications furthers the client’s intent.

Commentators on the law also recognize that the general rule is that the attorney-client privilege continues after death. See, e. g., 8 Wigmore, Evidence §2323 (MeNaughton rev. 1961); Frankel, The Attorney-Client Privilege After the Death of the Client, 6 Geo. J. Legal Ethics 45, 78-79 (1992); 1 J. Strong, McCormick on Evidence § 94, p. 348 (4th ed. 1992). Undoubtedly, as the Independent Counsel emphasizes, various commentators have criticized this rule, urging that the privilege should be abrogated after the client’s death where extreme injustice would result, as long as disclosure would not seriously undermine the privilege by deterring elient communication. See, e.g., C. Mueller & L. Kirkpatrick, 2 Federal Evidence §199, pp. 380-381 (2d ed. 1994); Restatement (Third) of the Law Governing Lawyers § 127, Comment *407d (Proposed Final Draft No. 1, Mar. 29, 1996). But even these critics clearly recognize that established law supports the continuation of the privilege and that a contrary rule would be a modification of the common law. See, e. g., Mueller & Kirkpatrick, supra, at 379; Restatement of the Law Governing Lawyers, supra, § 127, Comment c; 24 C. Wright & K. Graham, Federal Practice and Procedure § 5498, p. 488 (1986).

Despite the scholarly criticism, we think there are weighty reasons that counsel in favor of posthumous application. Knowing that communications will remain confidential even after death encourages the client to communicate fully and frankly with counsel. While the fear of disclosure, and the consequent withholding of information from counsel, may be reduced if disclosure is limited to posthumous disclosure in a criminal context, it seems unreasonable to assume that it vanishes altogether. Clients may be concerned about reputation, civil liability, or possible harm to friends or family. Posthumous disclosure of such communications may be as feared as disclosure during the client’s lifetime.

The Independent Counsel suggests, however, that his proposed exception would have little to no effect on the client’s willingness to confide in his attorney. He reasons that only clients intending to perjure themselves will be chilled by a ride of disclosure after death, as opposed to truthful clients or those asserting their Fifth Amendment privilege. This is because for the latter group, communications disclosed by the attorney after the client’s death purportedly will reveal only information that the client himself would have revealed if alive.

The Independent Counsel assumes, incorrectly we believe, that the privilege is analogous to the Fifth Amendment’s protection against self-incrimination. But as suggested above, the privilege serves much broader purposes. Clients consult attorneys for a wide variety of reasons, only one of which involves possible criminal liability. Many attorneys *408aet as counselors on personal and family matters, where, in the course of obtaining the desired advice, confidences about family members or financial problems must be revealed in order to assure sound legal advice. The same is true of owners of small businesses who may regularly consult their attorneys about a variety of problems arising in the course of the business. These confidences may not come close to any sort of admission of criminal wrongdoing, but nonetheless be matters which the client would not wish divulged.

The contention that the attorney is being required to disclose only what the client could have been required to disclose is at odds with the basis for the privilege even during the client’s lifetime. In related cases, we have said that the loss of evidence admittedly caused by the privilege is justified in part by the fact that without the privilege, the client may not have made such communications in the first place. See Jaffee, 518 U. S., at 12; Fisher v. United States, 425 U. S. 391, 403 (1976). This is true of disclosure before and after the client’s death. Without assurance of the privilege’s posthumous application, the client may very well not have made disclosures to his attorney at all, so the loss of evidence is more apparent than real. In the case at hand, it seems quite plausible that Foster, perhaps already contemplating suicide, may not have sought legal advice from Hamilton if he had not been assured the conversation was privileged.

The Independent Counsel additionally suggests that his proposed exception would have minimal impact if confined to criminal cases, or, as the Court of Appeals suggests, if it is limited to information of substantial importance to a particular criminal case.3 However, there is no case authority for the proposition that the privilege applies differently in erimi-*409nal and civil cases, and only one commentator ventures such a suggestion, see Mueller & Kirkpatrick, supra, at 380-381. In any event, a client may not know at the time he discloses information to his attorney whether it will later be relevant to a civil or a criminal matter, let alone whether it will be of substantial importance. Balancing ex post the importance of the information against client interests, even limited to criminal eases, introduces substantial uncertainty into the privilege’s application. For just that reason, we have rejected use of a balancing test in defining the contours of the privilege. See Upjohn, 449 U. S., at 393; Jaffee, supra, at 17-18.

In a similar vein, the Independent Counsel argues that existing exceptions to the privilege, such as the crime-fraud exception and the testamentary exception, make the impact of one more exception marginal. However, these exceptions do not demonstrate that the impact of a posthumous exception would be insignificant, and there is little empirical evidence on this point.4 The established exceptions are con*410sistent with the purposes of the privilege, see Glover, 165 U. S., at 407-408; United States v. Zolin, 491 U. S. 554, 562-563 (1989), while a posthumous exception in criminal cases appears at odds with the goals of encouraging fall and frank communication and of protecting the client’s interests. A “no harm in one more exception” rationale could contribute to the general erosion of the privilege, without reference to common-law principles or “reason and experience.”

_ Finally, the Independent Counsel, relying on cases such as United States v. Nixon, 418 U. S. 683, 710 (1974), and Branzburg v. Hayes, 408 U. S. 665 (1972), urges that privileges be strictly construed because they are inconsistent with the paramount judicial goal of truth seeking. But both Nixon and Bmnzburg dealt with the creation of privileges not recognized by the common law, whereas here we deal with one of the oldest recognized privileges in the law. And we are asked, not simply to “construe” the privilege, but to narrow it, contrary to the weight of the existing body of case law.

It has been generally, if not universally, accepted, for well over a century, that the attorney-client privilege survives the death of the client in a case such as this. While the arguments against the survival of the privilege are by no means frivolous, they are based in large part on speculation — thoughtful speculation, but speculation nonetheless— as to whether posthumous termination of the privilege would diminish a client’s willingness to confide in an attorney. In an area where empirical information would be useful, it is scant and inconclusive.

Rule 501’s direction to look to “the principles of the common law as they may be interpreted by the courts of the United States in the light of reason and experience” does not mandate that a rule, once established, should endure for all time. Funk v. United States, 290 U. S. 371, 381 (1933). But *411here the Independent Counsel has simply not made a sufficient showing to overturn the common-law rule embodied in the prevailing ease law. Interpreted in the light of reason and experience, that body of law requires that the attorney-client privilege prevent disclosure of the notes at issue in this case. The judgment of the Court of Appeals is

Reversed.

Notes

Because we sustain the claim of attorney-client privilege, we do not reach the claim of work-product privilege.

About half the States have codified the testamentary exception by providing that a personal representative of the deceased can waive the privilege when heirs or devisees claim through the deceased client (as opposed to parties claiming against the estate, for whom the privilege is not waived). See, e. g., Ala. Rule Evid. 502 (1996); Ark. Code Ann. § 16-41-101, Rule 502 (Supp. 1997); Neb. Rev. Stat. §27-503, Rule 503 (1995). These statutes do not address expressly the continuation of the privilege outside the context of testamentary disputes, although many allow the attorney to assert the privilege on behalf of the client apparently without temporal limit. See, e. g., Ark. Code Ann. § 16-41-101, Rule 502(e) (Supp. 1997). They thus do not refute or affirm the general presumption in the case law that the privilege survives. California’s statute is exceptional in that it apparently allows the attorney to assert the privilege only so long as a holder of the privilege (the estate’s personal representative) exists, suggesting the privilege terminates when the estate is wound up. See Cal. Code Evid. Ann. §§954, 957 (West 1995). But no other State has followed California’s lead in this regard.

Petitioners, while opposing wholesale abrogation of the privilege in criminal cases, concede that exceptional eireumstanees implicating a criminal defendant’s constitutional rights might warrant breaching the privilege. We do not, however, need to reach this issue, since such exceptional circumstances clearly are not presented here.

Empirical evidence on the privilege is limited. Three studies do not reach firm conclusions on whether limiting the privilege would discourage full and frank communication. Alexander, The Corporate Attorney Client Privilege: A Study of the Participants, 63 St. John’s L. Rev. 191 (1989); Zacharias, Rethinking Confidentiality, 74 Iowa L. Rev. 352 (1989); Comment, Functional Overlap Between the Lawyer and Other Professionals: Its Implications for the Privileged Communications Doctrine, 71 Yale L. J. 1226 (1962). These articles note that clients are often uninformed or mistaken about the privilege, but suggest that a substantial number of clients and attorneys think the privilege encourages candor. Two of the articles conclude that a substantial number of clients and attorneys think the privilege enhances open communication, Alexander, supra, at 244-246,261, and that the absence of a privilege would be detrimental to such communication, Comment, 71 Yale L. J., supra, at 1236. The third article suggests instead that while the privilege is perceived as important to open communication, limited exceptions to the privilege might not discourage such communication, Zacharias, supra, at 382, 386. Similarly, relatively few court decisions discuss the impact of the privilege’s application after death. This may reflect the general assumption that the privilege sur*410vives — if attorneys were required as a matter of practice to testify or provide notes in criminal proceedings, cases discussing that practice would surely exist.



Dissenting Opinion

Justice O’Connor,

with whom Justice Scalia and Justice Thomas join, dissenting.

Although the attorney-client privilege ordinarily will survive the death of the client, I do not agree with the Court that it inevitably precludes disclosure of a deceased client’s communications in criminal proceedings. In my view, a criminal defendant’s right to exculpatory evidence or a compelling law enforcement need for information may, where the testimony is not available from other sources, override a client’s posthumous interest in confidentiality.

We have long recognized that “[t]he fundamental basis upon which all rules of evidence must rest — if they are to rest upon reason — is their adaptation to the successful development of the truth.” Funk v. United States, 290 U. S. 371, 381 (1933). In light of the heavy burden that they place on the search for truth, see United States v. Nixon, 418 U. S. 683, 708-710 (1974), “[e]videntiary privileges in litigation are not favored, and even those rooted in the Constitution must give way in proper circumstances,” Herbert v. Lando, 441 U. S. 153, 175 (1979). Consequently, we construe the scope of privileges narrowly. See Jaffee v. Redmond, 518 U. S. 1, 19 (1996) (Scalia, J., dissenting); see also University of Pennsylvania v. EEOC, 493 U. S. 182, 189 (1990). We are reluctant to recognize a privilege or read an existing one expansively unless to do so will serve a “public good transcending the normally predominant principle of utilizing all rational means for ascertaining truth.” Trammel v. United *412States, 445 U. S. 40, 50 (1980) (internal quotation marks omitted).

The attorney-client privilege promotes trust in the representational relationship, thereby facilitating the provision of legal services and ultimately the administration of justice. See Upjohn Co. v. United States, 449 U. S. 383, 389 (1981). The systemic benefits of the privilege are commonly understood to outweigh the harm caused by excluding critical evidence. A privilege should operate, however, only where “necessary to achieve its purpose,” see Fisher v. United States, 425 U. S. 391, 403 (1976), and an invocation of the attorney-client privilege should not go unexamined “when it is shown that the interests of the administration of justice can only be frustrated by [its] exercise,” Cohen v. Jenkintown Cab Co., 238 Pa. Super. 456, 464, 357 A. 2d 689, 693-694 (1976).

I agree that a deceased client may retain a personal, repu-tational, and economic interest in confidentiality. See ante, at 407. But, after death, the potential that disclosure will harm the client's interests has been greatly diminished, and the risk that the client will be held criminally liable has abated altogether. Thus, some commentators suggest that terminating the privilege upon the client’s death “could not to any substantial degree lessen the encouragement for free disclosure which is [its] purpose.” 1 J. Strong, McCormick on Evidence § 94, p. 350 (4th ed. 1992); see also Restatement (Third) of the Law Governing Lawyers §127, Comment d (Proposed Final Draft No. 1, Mar. 29, 1996). This diminished risk is coupled with a heightened urgency for discovery of a deceased client's communications in the criminal context. The privilege does not “protect disclosure of the underlying facts by those who communicated with the attorney,” Upjohn, supra, at 395, and were the client living, prosecutors could grant immunity and compel the relevant testimony. After a client’s death, however, if the privilege precludes an attorney from testifying in the client’s stead, a complete *413"loss of crucial information” will often result, see 24 C. Wright & K. Graham, Federal Practice and Procedure § 5498, p. 484 (1986).

As the Court of Appeals observed, the costs of recognizing an absolute posthumous privilege can be inordinately high. See In re Sealed Case, 124 F. 3d 230, 233-234 (CADC 1997). Extreme injustice may occur, for example, where a criminal defendant seeks disclosure of a deceased client’s confession to the offense. See State v. Macumber, 112 Ariz. 569, 571, 544 P. 2d 1084, 1086 (1976); cf. In the Matter of a John Doe Grand Jury Investigation, 408 Mass. 480, 486, 562 N. E. 2d 69, 72 (1990) (Nolan, J., dissenting). In my view, the paramount value that our criminal justice system places on protecting an innocent defendant should outweigh a deceased client’s interest in preserving confidences. See, e. g., Schlup v. Delo, 513 U. S. 298, 324-325 (1995); In re Winship, 397 U. S. 358, 371 (1970) (Harlan, J., concurring). Indeed, even petitioners acknowledge that an exception may be appropriate where the constitutional rights of a criminal defendant are at stake. An exception may likewise be warranted in the face of a compelling law enforcement need for the information. “[0]ur historic commitment to the rule of law . . . is nowhere more profoundly manifest than in our view that the twofold aim of criminal justice is that guilt shall not escape or innocence suffer.” Nixon, supra, at 709 (internal quotation marks omitted); see also Herrera v. Collins, 506 U. S. 390, 398 (1993). Given that the complete exclusion of relevant evidence from a criminal trial or investigation may distort the record, mislead the factfinder, and undermine the central truth-seeking function of the courts, I do not believe that the attorney-client privilege should act as an absolute bar to the disclosure of a deceased client’s communications. When the privilege is asserted in the criminal context, and a showing is made that the communications at issue contain necessary factual information not otherwise available, courts *414should be permitted to assess whether interests in fairness and accuracy outweigh the justifications for the privilege.

A number of exceptions to the privilege already qualify its protections, and an attorney “who tells his client that the expected communications are absolutely and forever privileged is oversimplifying a bit.” 124 F. 3d, at 235. In the situation where the posthumous privilege most frequently arises — a dispute between heirs over the decedent's will— the privilege is widely recognized to give way to the interest in settling the estate. See Glover v. Patten, 165 U. S. 394, 406-408 (1897). This testamentary exception, moreover, may be invoked in some eases where the decedent would not have chosen to waive the privilege. For example, “a decedent might want to provide for an illegitimate child but at the same time much prefer that the relationship go undisclosed.” 124 F. 3d, at 234. Among the Court’s rationales for a broad construction of the posthumous privilege is its assertion that “[m]any attorneys act as counselors on personal and family matters, where, in the course of obtaining the desired advice, confidences about family members or financial problems must be revealed . . . which the client would not wish divulged.” Ante, at 407-408. That reasoning, however, would apply in the testamentary context with equal force. Nor are other existing exceptions to the privilege — for example, the crime-fraud exception or the exceptions for claims relating to attorney competence or compensation — necessarily consistent with “encouraging full and frank communication” or “protecting the client’s interests.” Ante, at 410. Rather, those exceptions reflect the understanding that, in certain circumstances, the privilege “ ‘ceases to operate’ ” as a safeguard on “the proper functioning of our adversary system.” See United States v. Zolin, 491 U. S. 554, 562-563 (1989).

Finally, the common law authority for the proposition that the privilege remains absolute after the client’s death is not a monolithic body of precedent. Indeed, the Court acknowl*415edges that most eases merely "presume the privilege survives,” see ante, at 404, and it relies on the case law’s "implicit acceptance” of a continuous privilege, see ante, at 406. Opinions squarely addressing the posthumous force of the privilege "are relatively rare.” See 124 F. 3d, at 232. And even in those decisions expressly holding that the privilege continues after the death of the client, courts do not typically engage in detailed reasoning, but rather conclude that the eases construing the testamentary exception imply survival of the privilege. See, e. g., Glover, supra, at 406-408; see also Wright & Graham, supra, §5498, at 484 (“Those who favor an eternal duration for the privilege seldom do much by way of justifying this in terms of policy”).

Moreover, as the Court concedes, see ante, at 403-404, 406-407, there is some authority for the proposition that a deceased client’s communications may be revealed, even in circumstances outside of the testamentary context. California’s Evidence Code, for example, provides that the attorney-client privilege continues only until the deceased client’s estate is finally distributed, noting that “there is little reason to preserve secrecy at the expense of excluding relevant evidence after the estate is wound up and the representative is discharged.” Cal. Evid. Code Ann. §954, and comment, p. 232, §952 (West 1995). And a state appellate court has admitted an attorney’s testimony concerning a deceased client’s communications after "balane[ing] the necessity for revealing the substance of the [attorney-client conversation] against the unlikelihood of any cognizable injury to the rights, interests, estate or memory of [the client].” See Cohen, supra, at 464, 357 A. 2d, at 693. The American Law Institute, moreover, has recently recommended withholding the privilege when the communication “bears on a litigated issue of pivotal significance” and has suggested that courts “balance the interest in confidentiality against any exceptional need for the communication.” Restatement (Third) of the Law Governing Lawyers § 127, at 431, Com*416ment d; see also 2 C. Mueller & L. Kirkpatrick, Federal Evidence, § 199, p. 380 (2d ed, 1994) (“[I]f a deceased client has confessed to criminal acts that are later charged to another, surely the latter’s need for evidence sometimes outweighs the interest in preserving the confidences”).

Where the exoneration of an innocent criminal defendant or a compelling law enforcement interest is at stake, the harm of precluding critical evidence that is unavailable by any other means outweighs the potential disincentive to forthright communication. In my view, the cost of silence warrants a narrow exception to the rule that the attorney-client privilege survives the death of the client. Moreover, although I disagree with the Court of Appeals’ notion that the context of an initial client interview affects the applicability of the work product doctrine, I do not believe that the doctrine applies where the material concerns a client who is no longer a potential party to adversarial litigation.

Accordingly, I would affirm the judgment of the Court of Appeals. Although the District Court examined the documents in camera, it has not had an opportunity to balance these competing considerations and decide whether the privilege should be trumped in the particular circumstances of this case. Thus, I agree with the Court of Appeals’ decision to remand for a determination whether any portion of the notes must be disclosed.

With respect, I dissent.