OCTOBER TERM, 1997
UNITED STATES v. BALSYS
CERTIORARI TO THE UNITED STATES COURT OF APPEALS FOR THE SECOND CIRCUIT
No. 97-873. Argued April 20, 1998-Decided June 25,1998
When the Office of Special Investigations of the Department of Justice's Criminal Division (OSI) subpoenaed respondent Balsys, a resident alien, to testify about his wartime activities between 1940 and 1944 and his immigration to the United States, he claimed the Fifth Amendment privilege against self-incrimination, based on his fear of prosecution by a foreign nation. The Federal District Court granted OSI's petition to enforce the subpoena, but the Second Circuit vacated the order, holding that a witness with a real and substantial fear of prosecution by a foreign country may assert the privilege to avoid giving testimony in a domestic proceeding, even if the witness has no valid fear of a criminal prosecution in this country.
Held: Concern with foreign prosecution is beyond the scope of the SelfIncrimination Clause. Pp. 671-700.
(a) As a resident alien, Balsys is a "person" who, under that Clause, cannot "be compelled in any criminal case to be a witness against himself." See Kwong Hai Ghew v. Golding, 344 U. S. 590 , 596. However, the question here is whether a criminal prosecution by a foreign government not subject to this country's constitutional guarantees presents a "criminal case" for purposes of the privilege. Pp. 671-672.
(b) Balsys initially relies on the textual contrast between the Sixth Amendment, which clearly applies only to domestic criminal proceedings, and the Fifth, with its broader reference to "any criminal case," to argue that "any criminal case" means exactly that, regardless of the prosecuting authority. But the argument overlooks the cardinal rule to construe provisions in context. See King v. St. Vincent's Hospital, 502 U. S. 215 , 221. Because none of the other provisions of the Fifth Amendment is implicated except by action of the government that it binds, it would have been strange to choose such associates for a Clause meant to take a broader view. Further, a more modest understanding, that "any criminal case" distinguishes the Fifth Amendment's SelfIncrimination Clause from its Clause limiting grand jury indictments to "capital, or otherwise infamous crime[s]," provides an explanation for the text of the privilege. Indeed, there is no known clear common-law precedent or practice, contemporaneous with the framing, for looking to
the possibility offoreign prosecution as a premise for claiming the privilege. Pp. 672-674.
(c) In the precursors of this case, the Court concluded that prosecution in a state jurisdiction not bound by the Self-Incrimination Clause is beyond the purview of the privilege. United States v. Murdock, 284 U. S. 141. United States v. Saline Bank of Va., 1 Pet. 100, and Ballmann v. Fagin, 200 U. S. 186 , distinguished. The Court's precedent turned away from this proposition once, in Malloy v. Hogan, 378 U. S. 1 , 3, where it applied the Fourteenth Amendment due process incorporation to the Self-Incrimination Clause, so as to bind the States as well as the National Government by its terms. It immediately said, in Murphy v. Waterfront Comm'n of N. Y. Harbor, 378 U. S. 52 , 57, that Malloy necessitated a reconsideration of Murdock's rule. After Malloy, the Fifth Amendment limitation was no longer framed for one jurisdiction alone, each jurisdiction having instead become subject to the same privilege claim flowing from the same source. Since fear of prosecution in the one jurisdiction now implicated the very privilege binding upon the other, the Murphy opinion sensibly recognized that if a witness could not assert the privilege in such circumstances, the witness could be "whipsawed" into incriminating himself under both state and federal law, even though the privilege was applicable to each. Such whipsawing is possible because the privilege against self-incrimination can be exchanged by the government for an immunity to prosecutorial use of any compelled inculpatory testimony. Kastigar v. United States, 406 U. S. 441, 448-449. Such an exchange by the government is permissible only when it provides immunity as broad as the privilege. Mter Malloy had held the privilege binding on the state jurisdictions as well as the National Government, it would have been intolerable to allow a prosecutor in one or the other jurisdiction to eliminate the privilege by offering immunity less complete than the privilege's dual jurisdictional reach. To the extent that the Murphy Court undercut Murdock's rationale on historical grounds, its reasoning that English cases supported a more expansive reading of the Clause is flawed and cannot be accepted now. Pp.674-690.
(d) Murphy discusses a catalog of "Policies of the Privilege," which could suggest a concern broad enough to encompass foreign prosecutions. However, the adoption of such a revised theory would rest on Murphy's treatment of English cases, which has been rejected as an indication of the Clause's meaning. Moreover, although Murphy catalogs aspirations furthered by the Clause, its discussion does not weigh the host of competing policy concerns that would be raised in a legitimate reconsideration of the Clause's scope. Contrary to Balsys's
contention, general personal testimonial integrity or privacy is not a reliable guide to the Clause's scope of protection. Fifth Amendment tradition offers, in practice, a conditional protection of testimonial privacy. Since the judiciary could not recognize fear of foreign prosecution and at the same time preserve the Government's existing rights to seek testimony in exchange for immunity (because domestic courts could not enforce the immunity abroad), extending the privilege would change the balance of private and governmental interests that has been accepted for as long as there has been Fifth Amendment doctrine. Balsys also argues that Murphy's policy catalog supports application of the privilege in order to prevent the Government from overreaching to facilitate foreign criminal prosecutions in a spirit of "cooperative internationalism." Murphy recognized "cooperative federalism"-the teamwork of state and national officials to fight interstate crime-but only to underscore the significance of the Court's holding that a federal court could no longer ignore fear of state prosecution when ruling on a privilege claim. Since in this case there is no counterpart to Malloy, imposing the Fifth Amendment beyond the National Government, there is no premise in Murphy for appealing to "cooperative internationalism" by analogy to "cooperative federalism." The analogy must, instead, be to the pre-Murphy era when the States were not bound by the privilege. Even if "cooperative federalism" and "cooperative internationalism" did support expanding the privilege's scope, Balsys has not shown that the likely costs and benefits justify such expansion. Cooperative conduct between the United States and foreign nations may one day develop to a point at which fear of foreign prosecution could be recognized under the Clause as traditionally understood, but Balsys has presented no interest rising to such a level of cooperative prosecution. Pp. 690-700.
119 F.3d 122 , reversed and remanded.
SOUTER, J., delivered the opinion of the Court, in which REHNQUIST, C. J., and STEVENS, O'CONNOR, and KENNEDY, JJ., joined, and in which SCALIA and THOMAS, JJ., joined as to Parts I, II, and III. STEVENS, J., filed a concurring opinion, post, p. 700. GINSBURG, J., filed a dissenting opinion, post, p. 701. BREYER, J., filed a dissenting opinion, in which GINSBURG, J., joined, post, p. 702.
Deputy Solicitor General Dreeben argued the cause for the United States. With him on the briefs were Solicitor General Waxman, Acting Assistant Attorney General Keeney, Barbara McDowell, and Joseph C. Wyderko.
Ivars Berzins argued the cause and filed a brief for respondent. *
JUSTICE SOUTER delivered the opinion of the Court.t
By administrative subpoena, the Office of Special Investigations of the Criminal Division of the United States Department of Justice (OSI) sought testimony from the respondent, Aloyzas Balsys, about his wartime activities between 1940 and 1944 and his immigration to the United States in 1961. Balsys declined to answer such questions, claiming the Fifth Amendment privilege against self-incrimination, based on his fear of prosecution by a foreign nation. We hold that concern with foreign prosecution is beyond the scope of the Self- Incrimination Clause.
Respondent Aloyzas Balsys is a resident alien living in Woodhaven, New York, having obtained admission to this country in 1961 under the Immigration and Nationality Act, 8 U. S. C. § 1201, on an immigrant visa and alien registration issued at the American Consulate in Liverpool. In his application, he said that he had served in the Lithuanian army between 1934 and 1940, and had lived in hiding in Plateliai, Lithuania, between 1940 and 1944. Balsys swore that the information was true, and signed a statement of understanding that if his application contained any false information or materially misleading statements, or concealed any material fact, he would be subject to criminal prosecution and deportation.
*Elizabeth Holtzman and Sanford Hausler filed a brief for the World Jewish Congress et al. as amici curiae urging reversal.
John D. Cline, Barbara E. Bergman, and John L. Pollok filed a brief for the National Association of Criminal Defense Lawyers et al. as amici curiae urging affirmance.
tJUSTICE SCALIA and JUSTICE THOMAS join only Parts I, II, and III of this opinion.
OSI, which was created to institute denaturalization and deportation proceedings against suspected Nazi war criminals, is now investigating whether, contrary to his representations, Balsys participated in Nazi persecution during World War II. Such activity would subject him to deportation for persecuting persons because of their race, religion, national origin, or political opinion under §§ 1182(a)(3)(E) and 1251(a)(4)(D), as well as for lying on his visa application under §§ 1182(a)(6)(C)(i) and 1251(a)(1)(A).
When OSI issued a subpoena requiring Balsys to testify at a deposition, he appeared and gave his name and address, but he refused to answer any other questions, such as those directed to his wartime activities in Europe between 19401945 and his immigration to the United States in 1961. In response to all such questions, Balsys invoked the Fifth Amendment privilege against compelled self-incrimination, claiming that his answers could subject him to criminal prosecution. He did not contend that he would incriminate himself under domestic law,l but claimed the privilege because his responses could subject him to criminal prosecution by Lithuania, Israel, and Germany.
OSI responded with a petition in Federal District Court to enforce the subpoena under § 1225(a). Although the District Court found that if Balsys were to provide the information requested, he would face a real and substantial danger of prosecution by Lithuania and Israel (but not by Germany), it granted OS1's enforcement petition and ordered Balsys to testify, treating the Fifth Amendment as inapplicable to a claim of incrimination solely under foreign law. 918 F. Supp. 588 (EDNY 1996). Balsys appealed, and the Court of Appeals for the Second Circuit vacated the District Court's order, holding that a witness with a real and substantial fear of prosecution by a foreign country may assert the Fifth Amendment privilege to avoid giving testimony in a domes-
1 The Government advises us that the statute of limitation bars criminal prosecution for any misrepresentation. Tr. of Oral Arg. 4.
tic proceeding, even if the witness has no valid fear of a criminal prosecution in this country. 119 F.3d 122 (1997). We granted certiorari, 522 U. S. 1072 (1998), to resolve a conflict among the Circuits on this issue 2 and now reverse.
The Self-Incrimination Clause of the Fifth Amendment provides that "[n]o person ... shall be compelled in any criminal case to be a witness against himself." U. S. Const., Arndt. 5. Resident aliens such as Balsys are considered "persons" for purposes of the Fifth Amendment and are entitled to the same protections under the Clause as citizens. See Kwong Hai Ghew v. Golding, 344 U. S. 590 , 596 (1953). The parties do not dispute that the Government seeks to "compel" testimony from Balsys that would make him "a witness against himself." The question is whether there is a risk that Balsys's testimony will be used in a proceeding that is a "criminal case."
Balsys agrees that the risk that his testimony might subject him to deportation is not a sufficient ground for asserting the privilege, given the civil character of a deportation proceeding. See INS v. Lopez-Mendoza, 468 U. S. 1032 , 1038-1039 (1984). If, however, Balsys could demonstrate
2 See United States v. Gecas, 120 F.3d 1419 (CAll 1997) (en bane) (holding that the privilege cannot be invoked based on fear of prosecution abroad); United States v. (Under Seal), 794 F.2d 920 (CA4) (same), cert. denied sub nom. Araneta v. United States, 479 U. S. 924 (1986); In re Parker, 411 F.2d 1067 (CAW 1969) (same), vacated as moot, 397 U. S. 96 (1970).
We have granted certiorari in cases raising this question twice before but did not reach its merits in either case. See Zicarelli v. New Jersey Comm'n of Investigation, 406 U. S. 472 (1972) (finding that because the petitioner did not face a "real and substantial" risk of foreign prosecution, it was unnecessary to decide whether the privilege can be asserted based on fear of foreign prosecution); Parker v. United States, 397 U. S. 96 (1970) (per curiam) (vacating and remanding with instructions to dismiss as moot).
that any testimony he might give in the deportation investigation could be used in a criminal proceeding against him brought by the Government of either the United States or one of the States, he would be entitled to invoke the privilege. It "can be asserted in any proceeding, civil or criminal, administrative or judicial, investigatory or adjudicatory," in which the witness reasonably believes that the information sought, or discoverable as a result of his testimony, could be used in a subsequent state or federal criminal proceeding. Kastigar v. United States, 406 U. S. 441 , 444-445 (1972); see also McCarthy v. Arndstein, 266 U. S. 34 , 40 (1924) (the privilege "applies alike to civil and criminal proceedings, wherever the answer might tend to subject to criminal responsibility him who gives it"). But Balsys makes no such claim, contending rather that his entitlement to invoke the privilege arises because of a real and substantial fear that his testimony could be used against him by Lithuania or Israel in a criminal prosecution. The reasonableness of his fear is not challenged by the Government, and we thus squarely face the question whether a criminal prosecution by a foreign government not subject to our constitutional guarantees presents a "criminal case" for purposes of the privilege against self-incrimination.
Balsys relies in the first instance on the textual contrast between the Sixth Amendment, which clearly applies only to domestic criminal proceedings, and the Compelled SelfIncrimination Clause, with its facially broader reference to "any criminal case." The same point is developed by Balsys's amici, 3 who argue that "any criminal case" means exactly that, regardless of the prosecuting authority. According to the argument, the Framers' use of the adjective "any" precludes recognition of the distinction raised by the
3 See Brief for National Association of Criminal Defense Lawyers et al. as Amici Curiae 5.
Government, between prosecution by a jurisdiction that is itself bound to recognize the privilege and prosecution by a foreign jurisdiction that is not. But the argument overlooks the cardinal rule to construe provisions in context. See King v. St. Vincent's Hospital, 502 U. S. 215 , 221 (1991). In the Fifth Amendment context, the Clause in question occurs in the company of guarantees of grand jury proceedings, defense against double jeopardy, due process, and compensation for property taking. Because none of these provisions is implicated except by action of the government that it binds, it would have been strange to choose such associates for a Clause meant to take a broader view, and it would be strange to find such a sweep in the Clause now. See Wharton v. Wise, 153 U. S. 155 , 169-170 (1894) (noscitur a sociis); see also Gustafson v. Alloyd Co., 513 U. S. 561 , 575 (1995) (same). The oddity of such a reading would be especially stark if the expansive language in question is open to another reasonable interpretation, as we think it is. Because the Fifth Amendment opens by requiring a grand jury indictment or presentment "for a capital, or otherwise infamous crime,"4 the phrase beginning with "any" in the subsequent Self- Incrimination Clause may sensibly be read as making it clear that the privilege it provides is not so categorically limited. It is plausible to suppose the adjective was inserted only for that purpose, not as taking the further step of defining the relevant prosecutorial jurisdiction internationally. We therefore take this to be the fair reading of the adjective "any," and we read the Clause contextually as
4 As a whole, the Amendment reads as follows: "No person shall be held to answer for a capital, or otherwise infamous crime, unless on a presentment or indictment of a Grand Jury, except in cases arising in the land or naval forces, or in the Militia, when in actual service in time of War or public danger; nor shall any person be subject for the same offence to be twice put in jeopardy of life or limb; nor shall be compelled in any criminal case to be a witness against himself, nor be deprived of life, liberty, or property, without due process of law; nor shall private property be taken for public use, without just compensation."
apparently providing a witness with the right against compelled self-incrimination when reasonably fearing prosecution by the government whose power the Clause limits, but not otherwise. Since there is no helpful legislative history,5 and because there was no different common law practice at the time of the framing, see Part III-C, infra; cf. Counselman v. Hitchcock, 142 U. S. 547 , 563-564 (1892) (listing a sample of cases, including preframing cases, in which the privilege was asserted, none of which involve fear of foreign prosecution), there is no reason to disregard the contextual reading. This Court's precedent has indeed adopted that so-called same-sovereign interpretation.
The currently received understanding of the Bill of Rights as instituted "to curtail and restrict the general powers granted to the Executive, Legislative, and Judicial Branches" of the National Government defined in the original constitutional articles, New York Times Co. v. United
5 See Gecas, 120 F. 3d, at 1435 (noting that the Clause has "virtually no legislative history"); 5 The Founders' Constitution 262 (P. Kurland & R. Lerner eds. 1987) (indicating that the Clause as originally drafted and introduced in the First Congress lacked the phrase "any criminal case," which was added at the behest of Representative Lawrence on the ground that the Clause would otherwise be "in some degree contrary to laws passed").
In recent years, scholarly attention has refined our knowledge of the previous manifestations of the privilege against self-incrimination, the present culmination of such scholarship being R. Helmholz et al., The Privilege Against Self-Incrimination (1997). What we know of the circumstances surrounding the adoption of the Fifth Amendment, however, gives no indication that the Framers had any sense of a privilege more comprehensive than common law practice then revealed. See Moglen, Taking the Fifth: Reconsidering the Origins of the Constitutional Privilege Against Self-Incrimination, 92 Mich. L. Rev. 1086, 1123 (1994) ("[T]he legislative history of the Fifth Amendment adds little to our understanding of the history of the privilege"). As to the common law practice, see Part III-C, infra.
States, 403 U. S. 713 , 716 (1971) (per curiam) (Black, J., concurring) (emphasis deleted), was expressed early on in Chief Justice Marshall's opinion for the Court in the leading case of Barron ex rel. Tiernan v. Mayor of Baltimore, 7 Pet. 243, 247 (1833): the Constitution's "limitations on power ... are naturally, and, we think, necessarily applicable to the government created by the instrument," and not to "distinct [state] governments, framed by different persons and for different purposes."
To be sure, it would have been logically possible to decide (as in Barron) that the "distinct [state] governments ... framed ... for different purposes" were beyond the ambit of the Fifth Amendment, and at the same time to hold that the self-incrimination privilege, good against the National Government, was implicated by fear of prosecution in another jurisdiction. But after Barron and before the era of Fourteenth Amendment incorporation, that would have been an unlikely doctrinal combination, and no such improbable development occurred.
The precursors of to day's case were those raising the question of the significance for the federal privilege of possible use of testimony in state prosecution. Only a handful of early cases even touched on the problem. In Brown v. Walker, 161 U. S. 591 (1896), a witness raised the issue, claiming the privilege in a federal proceeding based on his fear of prosecution by a State, but we found that a statute under which immunity from federal prosecution had been conferred provided for immunity from state prosecution as well, obviating any need to reach the issue raised. Id., at 606-608. In Jack v. Kansas, 199 U. S. 372 (1905), a Fourteenth Amendment case, we affirmed a sentence for contempt imposed on a witness in a state proceeding who had received immunity from state prosecution but refused to answer questions based on a fear that they would subject him to federal prosecution. Although there was no reasonable fear of a prosecution by the National Government in that
case, we addressed the question whether a self-incrimination privilege could be invoked in the one jurisdiction based on fear of prosecution by the other, saying that "[w]e think the legal immunity is in regard to a prosecution in the same jurisdiction, and when that is fully given it is enough." Id., at 382. A year later, in the course of considering whether a federal witness, immunized from federal prosecution, could invoke the privilege based on fear of state prosecution, we adopted the general proposition that "the possibility that information given by the witness might be used" by the other government is, as a matter of law, "a danger so unsubstantial and remote" that it fails to trigger the right to invoke the privilege. Hale v. Henkel, 201 U. S. 43 , 69 (1906).
"[I]f the argument were a sound one it might be carried still further and held to apply not only to state prosecutions within the same jurisdiction, but to prosecutions under the criminal laws of other States to which the witness might have subjected himself. The question has been fully considered in England, and the conclusion reached by the courts of that country [is] that the only danger to be considered is one arising within the same jurisdiction and under the same sovereignty. Queen v. Boyes, 1 B. & S. 311[, 121 Eng. Rep. 730]; King of the Two Sicilies v. Willcox, 7 State Trials (N. S.), 1049, 1068; State v. March, 1 Jones (N. Car.), 526; State v. Thomas, 98 N. Car. 599." Ibid.
A holding to this effect came when United States v. Murdock, 284 U. S. 141 (1931), "definitely settled" the question whether in a federal proceeding the privilege applied on account of fear of state prosecution, concluding "that one under examination in a federal tribunal could not refuse to answer on account of probable incrimination under state law." United States v. Murdock, 290 U. S. 389 , 396 (1933).
"The English rule of evidence against compulsory selfincrimination, on which historically that contained in
the Fifth Amendment rests, does not protect witnesses against disclosing offenses in violation of the laws of another country. King of the Two Sicilies v. Willcox, 7 State Trials (N. S.) 1049, 1068. Queen v. Boyes, 1 B. & S., at 330[, 121 Eng. Rep., at 738]. This court has held that immunity against state prosecution is not essential to the validity of federal statutes declaring that a witness shall not be excused from giving evidence on the ground that it will incriminate him, and also that the lack of state power to give witnesses protection against federal prosecution does not defeat a state immunity statute. The principle established is that full and complete immunity against prosecution by the government compelling the witness to answer is equivalent to the protection furnished by the rule against compulsory self-incrimination. Counselman v. Hitchcock, 142 U. S. 547 . Brown v. Walker, 161 U. S. 591 , 606. Jack v. Kansas, 199 U. S. 372 , 381. Hale v. Henkel, 201 U. S. 43 , 68. As appellee at the hearing did not invoke protection against federal prosecution, his plea is without merit and the government's demurrer should have been sustained." Murdock, 284 U. S., at 149.
Murdock's resolution of the question received a subsequent complement when we affirmed again that a State could compel a witness to give testimony that might incriminate him under federal law, see Knapp v. Schweitzer, 357 U. S. 371 (1958), overruled by Murphy v. Waterfront Comm'n of N. Y. Harbor, 378 U. S. 52 (1964), testimony that we had previously held to be admissible into evidence in the federal courts, see Feldman v. United States, 322 U. S. 487 (1944), overruled by Murphy, supra, at 80.
It has been suggested here that our precedent addressing fear of prosecution by a government other than the compelling authority fails to reflect the Murdock rule uniformly.
In 1927 (prior to our decision in Murdock), in a case involving a request for habeas relief from a deportation order, we declined to resolve whether "the Fifth Amendment guarantees immunity from self-incrimination under state statutes." United States ex rel. Vajtauer v. Commissioner of Immigration, 273 U. S. 103 , 113 (1927). Although we found that the witness had waived his claim to the privilege, our decision might be read to suggest that there was some tension between the reasoning of two of the cases discussed above, Hale v. Henkel and Brown v. Walker, and the analyses contained in two others, United States v. Saline Bank of Va., 1 Pet. 100 (1828), and BaUmann v. Fagin, 200 U. S. 186 (1906). 273 U. S., at 113. These last two cases have in fact been cited here for the claim that prior to due process incorporation, the privilege could be asserted in a federal proceeding based on fear of prosecution by a State.6 Saline Bank and BaUmann are not, however, inconsistent with Murdock.
In Saline Bank, we permitted the defendants to refuse discovery sought by the United States in federal court, where the defendants claimed that their responses would result in incrimination under the laws of Virginia. "The rule clearly is, that a party is not bound to make any discovery which would expose him to penalties, and this case falls within it." 1 Pet., at 104. But, for all the sweep of this statement, the opinion makes no mention of the Fifth Amendment, and in Hale v. Henkel, we explained that "the prosecution [in Saline Bank] was under a state law which imposed the penalty, and ... the Federal court was simply
6 The language in Vajtauer that has been cited in support of this suggestion says only that our conclusion that the witness waived his claim of privilege "makes it unnecessary for us to consider the extent to which the Fifth Amendment guarantees immunity from self-incrimination under state statutes or whether this case is to be controlled by Hale v. Henkel, 201 U. S. 43 ; Brown v. Walker, 161 U. S. 591 , 608; compare United States v. Saline Bank, 1 Pet. 100; BaUmann v. Fagin, 200 U. S. 186 , 195." 273 U. S., at 113.
administering the state law." 201 U. S., at 69. The state law, which addresses prosecutions brought by the State, suggested the rule that the Saline Bank Court applied to the case before it; the law provided that "no disclosure made by any party defendant to such suit in equity, and no books or papers exhibited by him in answer to the bill, or under the order of the Court, shall be used as evidence against him in any ... prosecution under this law," quoted in 1 Pet., at 104. Saline Bank, then, may have turned on a reading of state statutory law. Cf. McNaughton, Self-Incrimination Under Foreign Law, 45 Va. L. Rev. 1299, 1305-1306 (1959) (suggesting that Saline Bank represents "an application not of the privilege against self-incrimination ... but of the principle that equity will not aid a forfeiture"). But see BaUmann, supra, at 195 (Holmes, J.) (suggesting that Saline Bank is a Fifth Amendment case, though this view was soon repudiated by the Court in Hale, as just noted).
Where Saline Bank is laconic, BaUmann is equivocal.
While Ballmann specifically argued only the danger of incriminating himself under state law as his basis for invoking the privilege in a federal proceeding, and we upheld his claim of privilege, our opinion indicates that we concluded that Ballmann might have had a fear of incrimination under federal law as well as under state law. While we did suggest, contrary to the Murdock rule, that Ballmann might have been able to invoke the privilege based on a fear of state prosecution, the opinion says only that "[o]ne way or the other [due to the risk of incrimination under federal or state law] we are of opinion that Ballmann could not be required to produce his cash book if he set up that it would tend to criminate him." 200 U. S., at 195-196. At its equivocal worst, BaUmann reigned for only two months. Hale v. Henkel explained that "the only danger to be considered is one arising within the same jurisdiction and under the same sovereignty," 201 U. S., at 69, and BaUmann and Saline
Bank were later, of course, superseded by Murdock with its unequivocal holding that prosecution in a state jurisdiction not bound by the Clause is beyond the purview of the privilege.
In 1964, our precedent took a turn away from the unqualified proposition that fear of prosecution outside the jurisdiction seeking to compel testimony did not implicate a Fifth or Fourteenth Amendment privilege, as the case might be. In Murphy v. Waterfront Comm'n of N. Y. Harbor, 378 U. S. 52 (1964), we reconsidered the converse of the situation in Murdock, whether a witness in a state proceeding who had been granted immunity from state prosecution could invoke the privilege based on fear of prosecution on federal charges. In the course of enquiring into a work stoppage at several New Jersey piers, the Waterfront Commission of New York Harbor subpoenaed the defendants, who were given immunity from prosecution under the laws of New Jersey and New York. When the witnesses persisted in refusing to testify based on their fear of federal prosecution, they were held in civil contempt, and the order was affirmed by New Jersey's highest court. In re Application of the Waterfront Comm'n of N. Y. Harbor, 39 N. J. 436, 449, 189 A. 2d 36, 44 (1963). This Court held the defendants could be forced to testify not because fear of federal prosecution was irrelevant but because the Self-Incrimination Clause barred the National Government from using their state testimony or its fruits to obtain a federal conviction. We explained that "the constitutional privilege against self-incrimination protects a state witness against incrimination under federal as well as state law and a federal witness against incrimination under state as well as federal law." 378 U. S., at 77-78.
Murphy is a case invested with two alternative rationales.
Under the first, the result reached in Murphy was undoubtedly correct, given the decision rendered that very same day in Malloy v. Hogan, 378 U. S. 1 (1964), which applied the
doctrine of Fourteenth Amendment due process incorporation to the Self-Incrimination Clause, so as to bind the States as well as the National Government to recognize the privilege. Id., at 3. Prior to Malloy, the Court had refused to impose the privilege against self-incrimination against the States through the Fourteenth Amendment, see Twining v. New Jersey, 211 U. S. 78 (1908), thus leaving state-court witnesses seeking exemption from compulsion to testify to their rights under state law, as supplemented by the Fourteenth Amendment's limitations on coerced confessions. Malloy, however, established that "[t]he Fourteenth Amendment secures against state invasion the same privilege that the Fifth Amendment guarantees against federal infringement-the right of a person to remain silent unless he chooses to speak in the unfettered exercise of his own will, and to suffer no penalty ... for such silence." 378 U. S., at 8.
As the Court immediately thereafter said in Murphy, Malloy "necessitate[d] a reconsideration" of the unqualified Mur dock rule that a witness subject to testimonial compulsion in one jurisdiction, state or federal, could not plead fear of prosecution in the other. 378 U. S., at 57. After Malloy, the Fifth Amendment limitation could no longer be seen as framed for one jurisdiction alone, each jurisdiction having instead become subject to the same claim of privilege flowing from the one limitation. Since fear of prosecution in the one jurisdiction bound by the Clause now implicated the very privilege binding upon the other, the Murphy opinion sensibly recognized that if a witness could not assert the privilege in such circumstances, the witness could be "whipsawed into incriminating himself under both state and federal law even though the constitutional privilege against self-incrimination is applicable to each." 378 U. S., at 55 (internal quotation marks omitted).7 The whipsawing was possible owing to a
7 Prior to Murphy, such "whipsawing" efforts had been permissible, but arguably less outrageous since, as the opinion notes, "either the 'compelling' government or the 'using' government [was] a State, and, until today,
feature unique to the guarantee against self-incrimination among the several Fifth Amendment privileges. In the absence of waiver, the other such guarantees are purely and simply binding on the government. But under the SelfIncrimination Clause, the government has an option to exchange the stated privilege for an immunity to prosecutorial use of any compelled inculpatory testimony. Kastigar v. United States, 406 U. S., at 448-449. The only condition on the government when it decides to offer immunity in place of the privilege to stay silent is the requirement to provide an immunity as broad as the privilege itself. Id., at 449. After Malloy had held the privilege binding on the state jurisdictions as well as the National Government, it would therefore have been intolerable to allow a prosecutor in one or the other jurisdiction to eliminate the privilege by offering immunity less complete than the privilege's dual jurisdictional reach. Murphy accordingly held that a federal court could not receive testimony compelled by a State in the absence of a statute effectively providing for federal immunity, and it did this by imposing an exclusionary rule prohibiting the National Government "from making any such use of compelled testimony and its fruits," 378 U. S., at 79 (footnote omitted).
This view of Murphy as necessitated by Malloy was adopted in the subsequent case of Kastigar v. United States, supra, at 456, n. 42 ("Reconsideration of the rule that the Fifth Amendment privilege does not protect a witness in one jurisdiction against being compelled to give testimony that could be used to convict him in another jurisdiction was made necessary by the decision in Malloy v. Hogan"). Read this way, Murphy rests upon the same understanding of the Self- Incrimination Clause that Murdock recognized and to which the earlier cases had pointed. Although the Clause serves a variety of interests in one degree or another, see
the States were not deemed fully bound by the privilege against selfincrimination." 378 U. S., at 57, n. 6.
Part IV, infra, at its heart lies the principle that the courts of a government from which a witness may reasonably fear prosecution may not in fairness compel the witness to furnish testimonial evidence that may be used to prove his guilt. After Murphy, the immunity option open to the Executive Branch could be exercised only on the understanding that the state and federal jurisdictions were as one, with a federally mandated exclusionary rule filling the space between the limits of state immunity statutes and the scope of the privilege.8 As so understood, Murphy stands at odds with Balsys's claim.
There is, however, a competing rationale in Murphy, investing the Clause with a more expansive promise. The Murphy majority opened the door to this view by rejecting this Court's previous understanding of the English common-law evidentiary privilege against compelled selfincrimination, which could have informed the Framers' understanding of the Fifth Amendment privilege. See, e. g., Murphy, 378 U. S., at 67 (rejecting Murdock's analysis of the scope of the privilege under English common law). Having removed what it saw as an unjustified, historically derived
8 Of course, the judicial exclusion of compelled testimony functions as a fail-safe to ensure that compelled testimony is not admitted in a criminal proceeding. The general rule requires a grant of immunity prior to the compelling of any testimony. We have said that the prediction that a court in a future criminal prosecution would be obligated to protect against the evidentiary use of compelled testimony is not enough to satisfy the privilege against compelled self-incrimination. Pillsbury Co. v. Conboy, 459 U. S. 248 , 261 (1983). The suggestion that a witness should rely on a subsequent motion to suppress rather than a prior grant of immunity "would [not] afford adequate protection. Without something more, [the witness] would be compelled to surrender the very protection which the privilege is designed to guarantee." Maness v. Meyers, 419 U. S. 449 , 462 (1975) (footnote and internal quotation marks omitted). This general rule ensures that we do not "let the cat out with no assurance whatever of putting it back," id., at 463 (internal quotation marks omitted), and leaves the decision whether to grant immunity to the Executive in accord with congressional policy, see Pillsbury, supra, at 262.
limitation on the privilege, the Murphy Court expressed a comparatively ambitious conceptualization of personal privacy underlying the Clause, one capable of supporting, if not demanding, the scope of protection that Balsys claims. As the Court of Appeals recognized, if we take the Murphy opinion at face value, the expansive rationale can be claimed quite as legitimately as the Murdock-Malloy-Kastigar understanding of Murphy's result, and Balsys's claim accordingly requires us to decide whether Murphy's innovative side is as sound as its traditional one. We conclude that it is not.
As support for the view that the Court had previously misunderstood the English rule, Murphy relied, first, on two preconstitutional English cases, East India Co. v. Campbell, 1 Yes. sen. 246, 27 Eng. Rep. 1010 (Ex. 1749), and Brownsword v. Edwards, 2 Yes. sen. 243, 28 Eng. Rep. 157 (Ch. 1750), for the proposition that a witness in an English court was permitted to invoke the privilege based on fear of prosecution in a foreign jurisdiction. See 378 U. S., at 5859. Neither of these cases is on point as holding that proposition, however. In East India Co., a defendant before the Court of Exchequer, seeking to avoid giving an explanation for his possession of certain goods, claimed the privilege on the ground that his testimony might subject him to a fine or corporal punishment. The Court of Exchequer found that the defendant would be punishable in Calcutta, then an English Colony, and said it would "not oblige one to discover that, which, if he answers in the affirmative, will subject him to the punishment of a crime." 1 Yes. sen., at 247, 27 Eng. Rep., at 1011. In Brownsword, a defendant before the Court of Chancery claimed the privilege on the ground that her testimony could render her liable to prosecution in an English ecclesiastical court. "The general rule," the court said, "is that no one is bound to answer so as to subject himself to punishment, whether that punishment arises by the ecclesiastical law of the land." 2 Yes. sen., at 245, 28 Eng.
Rep., at 158. Although this statement, like its counterpart in East India Co., is unqualified, neither case is authority for the proposition that fear of prosecution in foreign courts implicates the privilege. For in each of these cases, the judicial system to which the witness's fears related was subject to the same legislative sovereignty that had created the courts in which the privilege was claimed.9 In fact, when these cases were decided, and for years after adoption of the Fifth Amendment, English authority was silent on whether fear of prosecution by a foreign nation implicated the privilege, and the Vice-Chancellor so stated in 1851. See King of the Two Sicilies v. Willcox, 1 Sim. (N. S.) 301, 331, 61 Eng. Rep. 116, 128 (Ch. 1851) (observing, in the course of an opinion that clearly involved a claim of privilege based on the fear of prosecution by another sovereign, that there is an "absence of all authority on the point").
Murphy, in fact, went on to discuss the case last cited, as well as a subsequent one. The Murphy majority began by acknowledging that King of the Two Sicilies was not authority for attacking this Court's prior view of English law. 378 U. S., at 60. In an opinion by Lord Cranworth, the Court of Chancery declined to allow defendants to assert the privilege
9 Further, the courts of both jurisdictions, at least in some cases, recognized the privilege against self-incrimination. East India Co. makes specific reference to the fact that the witness's testimony might be incriminating under the laws of Calcutta. 1 Yes. sen., at 247, 27 Eng. Rep., at 1011 ("[T]hat he is punishable appears from the case of Omichund v. Barker [1 Atk. 21, 26 Eng. Rep. 15 (1744)], as a jurisdiction is erected in Calcutta for criminal facts"). As of 1726, Calcutta was a "presidency town," which was subject to the civil jurisdiction of a "mayor's court." The mayor's court followed the English Rules of Evidence, which would have included the rule against self-incrimination. 1 Woodroffe & Ameer Ali's Law of Evidence in India 13 (P. Ramaswami & S. Rajagopalan eds., 11th ed. 1962). The ecclesiastical courts of England also recognized something akin to the privilege at this time in some cases. See Helmholz, Origins of the Privilege Against Self-Incrimination: The Role of the European Ius Commune, 65 N. Y. U. L. Rev. 962, 969-974 (1990) (citing cases heard in ecclesiastical courts in which the privilege was recognized).
based on their fear of prosecution in Sicily, for two reasons. 1 Sim. (N. S.), at 329, 61 Eng. Rep., at 128. The first was the court's belief that the privilege speaks only to matters that might be criminal under the laws of England: "The rule relied on by the Defendants, is one which exists merely by virtue of our own municipal law, and must, I think, have reference, exclusively, to matters penal by that law: to matters as to which, if disclosed, the Judge would be able to say, as matter of law, whether it could or could not entail penal consequences." For the second, the court relied on the unlikelihood that the defendants would ever leave England and be subject to Sicilian prosecution.
The Murphy majority nonetheless understood this rule to have been undermined by the subsequent case of United States of America v. McRae, 3 L. R. Ch. 79 (1867). See 378 U. S., at 61. In that suit brought by the United States against McRae in England to recover funds that he had collected there as a Confederate agent during the Civil War, the court recognized the privilege based on McRae's claim that his testimony would incriminate him in the United States. The court distinguished the litigation then before it from King of the Two Sicilies, indicating that though it agreed with the general principles stated by Lord Cranworth, see 3 L. R. Ch., at 84, he had not needed to lay down the broad proposition that invocation of the privilege was appropriate only with regard to matters penal under England's own law, see id., at 85. The court did not say that the privilege could be invoked in any case involving fear of prosecution under foreign law, however. Instead it noted two distinctions from King of the Two Sicilies, the first being that the "presumed ignorance of the Judge as to foreign law" on which King of the Two Sicilies rested has been "completely removed by the admitted statements upon the pleadings," 3 L. R. Ch., at 85; the second being that McRae presented the unusual circumstance that the party seeking to compel the testimony, the United States, was also the party
that would prosecute any crime under its laws that might thereby be revealed, id., at 87. The court's holding that the privilege could be invoked in such circumstances does not, however, support a general application of the privilege in any case in which a witness fears prosecution under foreign law by a party not before the court. Thus, Murphy went too far in saying that McRae overruled King of the Two Sicilies. lO See Murphy, 378 U. S., at 71. What is of more fundamental importance, however, is that even if McRae had announced a new development in English law going to the heart of King of the Two Sicilies, it would have been irrelevant to Fifth Amendment interpretation. The presumed influence of English law on the intentions of the Framers hardly invests the Framers with clairvoyance, and subsequent English developments are not attributable to the Framers by some rule of renvoi. Cf. Brown, 161 U. S., at 600 (citing Cathcart v. Robinson, 5 Pet. 264, 280 (1831)). Since McRae neither stated nor implied any disagreement with Lord Cranworth's 1857 statement in King of the Two Sicilies that there was no clear prior authority on the question, the Murphy Court had no authority showing that Murdock rested on unsound historical assumptions contradicted by opinions of the English courts.
10 Murphy also cites Heriz v. Riera, 11 Sim. 318, 59 Eng. Rep. 896 (1840), as support for the claim that the English rule allowed invocation of the privilege based on fear of prosecution abroad. See 378 U. S., at 63. In that case two Spanish women brought suit in England alleging that the defendant had violated a contract that he entered into with their brother and to which they were entitled to the proceeds as his heirs. The contract provided that the plaintiffs' brother (and they as his heirs) were entitled to a share of the proceeds from a mercantile contract with the Spanish Government. The defendant responded that the contract was illegal under the laws of Spain and hence unenforceable and resisted discovery because his answers might incriminate him under the Spanish code. The court accepted the defendant's plea, though it is unclear whether the court ruled on the merits of the plaintiffs' claim or the self-incrimination issue. See Grant, Federalism and Self-Incrimination, 5 UCLA L. Rev. 1,2 (1958).
In sum, to the extent that the Murphy majority went beyond its response to Malloy and undercut Murdock's rationale on historical grounds, its reasoning cannot be accepted now. Long before today, indeed, Murphy's history was shown to be fatally flawed.ll
11 Murphy, 378 U. S., at 81, n. 1 (Harlan, J., concurring in judgment) ("The English rule is not clear"); United States v. (Under Seal), 794 F. 2d, at 927 ("The Court's scholarship with respect to English law in this regard has been attacked, see Note, 69 Va. L. Rev. at 893-94 .... We do not enter the dispute as to whether Murphy represents a correct statement of the English rule at a particular time because we do not think that the Murphy holding depended upon the correctness of the Court's understanding of the state of English law and reliance thereon as the sole basis for decision. Rather, Murphy proceeds as a logical consequence to the holding in Malloy v. Hogan ... "); Note, Fifth Amendment Privilege Against Self-Incrimination and Fear of Foreign Prosecution, 96 Colum. L. Rev. 1940, 1944-1946, 1949, and nn. 79-81 (1996) ("The uncertainty of English law on [the question whether the privilege can be invoked based on fear of prosecution] casts doubt on the Supreme Court's holding in Murphy, which was based on the assertion that McRae 'represents the settled "English rule" regarding self-incrimination under foreign law.' Indeed, the Murphy Court's reliance on its idea of the 'true' English rule has been criticized by commentators, and its reading of British law was essentially overruled by the British Parliament. Murphy's reliance on mistaken interpretation and application of English law weakens its precedential value" (footnotes omitted)); Note, The Reach of the Fifth Amendment Privilege When Domestically Compelled Testimony May Be Used in a Foreign Country's Court, 69 Va. L. Rev. 875, 893-895 (1983) ("[T]he English rule argument has three fatal flaws. First, the so-called English rule, decided in 1867, never was the English rule despite overstatements by several American commentators and the Murphy Court. British commentators remained uncertain for nearly a century about the extent to which, if at all, their privilege protected against foreign incrimination .... Second, the English courts had not decided a case involving incrimination under the criminal laws of independent foreign sovereigns by the time our Constitution was framed. The only English cases involving independent sovereigns were decided more than sixty years later. Thus, even if the fifth amendment embodied the English common law at the time it was framed, the privilege did not incorporate any rule concerning foreign incrimination. Finally, even if the English rule protected against foreign incrimination, the Supreme Court in Zicarelli indicated that it had not
Although the Court and JUSTICE BREYER'S dissent differ on details, including some considerations of policy addressed in Part IV, infra, our basic disagreement with that dissent turns on three points. First, we start with what we think is the most probable reading of the Clause in its Fifth Amendment context, as limiting its principle to concern with prosecution by a sovereign that is itself bound by the Clause; the dissent instead emphasizes the Clause's facial breadth as consistent with a broader principle. Second, we rely on the force of our precedent, notably Murdock, as confirming this same-sovereign principle, as adapted to reflect the post Malloy requirement of immunity effective against both sovereigns subject to the one privilege under the National Constitution; the dissent attributes less force to Murdock, giving weight to its tension with the Saline Bank language, among other things. Third, we reject Murphy's restatement of the common-law background and read none of the common-law cases as authority inconsistent with our contextual reading of the Clause, later confirmed by precedent such as Murdock; the dissent finds support in the common-law cases for Murphy's historical reexamination and the broader reading of the Clause. In the end, our contextual reading of the Clause, combined with the Murdock holding, places a burden on any-
formally adopted the rule in Murphy" (footnotes omitted)); Capra, The Fifth Amendment and the Risk of Foreign Prosecution, N. Y. L. J., Mar. 8, 1991, p. 3 ("[D]espite Justice Goldberg's assertions in Murphy, it is clear that there was never a 'true' or uniform English rule .... [T]o the extent that the English rule would be pertinent to the Fifth Amendment privilege, it would have had to exist at the time the Fifth Amendment was adopted. Yet, as even Justice Goldberg admitted in Murphy, the English cases involving independent sovereigns were decided more than 60 years after the Fifth Amendment was adopted"); see also Law Reform Committee, Sixteenth Report, 1967, Cmnd. 3472, U1, p. 7 (explaining that English common law on the question is not "wholly consistent").
Murphy's reexamination of history also adopted the illegitimate reading of Saline Bank, rejected supra, at 678-679.
one who contests the basic same-sovereign principle, a burden that only clear, contrary, preframing common law might carry; since the dissent starts with a broader reading of the Clause and a less potent view of Murdock, it does not require Murphy and the common-law cases to satisfy such a burden before definitively finding that a more expansive principle underlies the Clause.
There remains, at least on the face of the Murphy majority's opinion, a further invitation to revise the principle of the Clause from what Murdock recognized. The Murphy majority opens its discussion with a catalog of "Policies of the Privilege," 378 U. S., at 55 (citations and internal quotation marks omitted):
"It reflects many of our fundamental values and most noble aspirations: our unwillingness to subject those suspected of crime to the cruel trilemma of selfaccusation, perjury or contempt; our preference for an accusatorial rather than an inquisitorial system of criminal justice; our fear that self-incriminating statements will be elicited by inhumane treatment and abuses; our sense of fair play which dictates a fair state-individual balance by requiring the government to leave the individual alone until good cause is shown for disturbing him and by requiring the government in its contest with the individual to shoulder the entire load; our respect for the inviolability of the human personality and of the right of each individual to a private enclave where he may lead a private life, our distrust of self-deprecatory statements; and our realization that the privilege, while sometimes a shelter to the guilty, is often a protection to the innocent."
Some of the policies listed would seem to point no further than domestic arrangements and so raise no basis for any privilege looking beyond fear of domestic prosecution. Oth-
ers, however, might suggest a concern broad enough to encompass foreign prosecutions and accordingly to support a more expansive theory of the privilege than the Murdock understanding would allow.
The adoption of any such revised theory would, however, necessarily rest on Murphy's reading of preconstitutional common-law cases as support for (or at least as opening the door to) the expansive view of the Framers' intent, which we and the commentators since Murphy have found to be unsupported. Once the Murphy majority's treatment of the English cases is rejected as an indication of the meaning intended for the Clause, Murdock must be seen as precedent at odds with Balsys's claim. That precedent aside, however, we think there would be sound reasons to stop short of resting an expansion of the Clause's scope on the highly general statements of policy expressed in the foregoing quotation from Murphy. While its list does indeed catalog aspirations furthered by the Clause, its discussion does not even purport to weigh the host of competing policy concerns that would be raised in a legitimate reconsideration of the Clause's scope.
The most general of Murphy's policy items ostensibly suggesting protection as comprehensive as that sought by Balsys is listed in the opinion as "the inviolability of the human personality and ... the right of each individual to a private enclave where he may lead a private life." 378 U. S., at 55 (internal quotation marks omitted). Whatever else those terms might cover, protection of personal inviolability and the privacy of a testimonial enclave would necessarily seem to include protection against the Government's very intrusion through involuntary interrogation.12 If in fact
12We are assuming, arguendo, that the intrusion is a subject of the Clause's protection. See Murphy, 378 U. S., at 57, n. 6; Gecas, 120 F. 3d, at 1462 (Birch, J., dissenting); cf. United States v. Verdugo-Urquidez, 494 U. S. 259, 264 (1990) ("The privilege against self-incrimination guaranteed
these values were reliable guides to the actual scope of protection under the Clause, they would be seen to demand a very high degree of protection indeed: "inviolability" is, after all, an uncompromising term, and we know as well from Fourth Amendment law as from a layman's common sense that breaches of privacy are complete at the moment of illicit intrusion, whatever use mayor may not later be made of their fruits. See United States v. Verdugo-Urquidez, 494 U. S. 259, 264 (1990) (citing United States v. Calandra, 414 U. S. 338 , 354 (1974); United States v. Leon, 468 U. S. 897 , 906 (1984)).
The Fifth Amendment tradition, however, offers no such degree of protection. If the Government is ready to provide the requisite use and derivative use immunity, see Kastigar, 406 U. S., at 453; see also Lefkowitz v. Turley, 414 U. S. 70 , 84 (1973), the protection goes no further: no violation of personality is recognized and no claim of privilege will availY One might reply that the choice of the word "inviolability" was just unfortunate; while testimonial integrity may not be inviolable, it is sufficiently served by requiring the Government to pay a price in the form of use (and derivative use) immunity before a refusal to testify will be overruled. But that answer overlooks the fact that when a witness's response will raise no fear of criminal penalty, there is no protection for testimonial privacy at all. See United States v. Ward, 448 U. S. 242 , 248-255 (1980).
Thus, what we find in practice is not the protection of personal testimonial inviolability, but a conditional protection of testimonial privacy subject to basic limits recognized before
by the Fifth Amendment is a fundamental trial right of criminal defendants. Although conduct by law enforcement officials prior to trial may ultimately impair that right, a constitutional violation occurs only at trial" (citation omitted)).
13 The practice of exchanging silence for immunity is unchallenged here and presumably invulnerable, being apparently as old as the Fifth Amendment itself. See Kastigar, 406 U. S., at 445, and n. 13.
the framing 14 and refined through immunity doctrine in the intervening years. Since the Judiciary could not recognize fear of foreign prosecution and at the same time preserve the Government's existing rights to seek testimony in exchange for immunity (because domestic courts could not enforce the immunity abroad), it follows that extending protection as Balsys requests would change the balance of private and governmental interests that has seemingly been accepted for as long as there has been Fifth Amendment doctrine. The upshot is that accepting personal testimonial integrity or privacy as a prima facie justification for the development Balsys seeks would threaten a significant change in the scope of traditional domestic protection; to the extent, on the other hand, that the domestic tradition is thought worthy of preservation, an appeal to a general personal testimonial integrity or privacy is not helpful. See Doe v. United States, 487 U. S. 201 , 213, n. 11 (1988) (finding no violation of the privilege "[d]espite the impact upon the inviolability of the human personality"); Schmerber v. California, 384 U. S. 757 , 762 (1966) (holding that a witness cannot rely on the privilege to decline to provide blood samples); ibid. ("[T]he privilege has never been given the full scope which the values that it helps to protect suggest").
Murphy's policy catalog would provide support, at a rather more concrete level, for Balsys's argument that application of the privilege in situations like his would promote the purpose of preventing government overreaching, which on anyone's view lies at the core of the Clause's purposes. This argument begins with the premise that "cooperative internationalism" creates new incentives for the Government to facilitate foreign criminal prosecutions. Because crime, like legitimate trade, is increasingly international, a correspond-
14 See n. 13, supra.
ing degree of international cooperation is coming to characterize the enterprise of criminal prosecution.15 The mission of the OS1 as shown in this case exemplifies the international cooperation that is said to undermine the legitimacy of treating separate governmental authorities as separate for purposes of liberty protection in domestic courts. Because the Government now has a significant interest in seeing individuals convicted abroad for their crimes, it is subject to the same incentive to overreach that has required application of the privilege in the domestic context. Balsys says that this argument is nothing more than the reasoning of the Murphy Court when it justified its recognition of a fear of state prosecution by looking to the significance of" 'cooperative federalism,'" the teamwork of state and national officials to fight interstate crime. 378 U. S., at 55-56.
But Balsys invests Murphy's "cooperative federalism" with a significance unsupported by that opinion. We have already pointed out that Murphy's expansion upon Murdock is not supported by Murphy's unsound historical reexamination, but must rest on Murphy's other rationale, under which its holding is a consequence of Malloy. That latter reading is essential to an understanding of "cooperative federalism." For the Murphy majority, "cooperative federalism" was not important standing alone, but simply because it underscored the significance of the Court's holding that after Malloy it would be unjustifiably formalistic for a federal court to ignore fear of state prosecution when ruling on a privilege claim. Thus, the Court described the "whipsaw" effect that the decision in Malloy would have created if fear of state prosecution were not cognizable in a federal proceeding:
"[The] policies and purposes [of the privilege] are defeated when a witness can be whipsawed into incriminating himself under both state and federal law
15 The Court of Appeals cited a considerable number of studies in the growing literature on the subject. 119 F.3d 122 , 130-131 (CA2 1997).
even though the constitutional privilege against selfincrimination is applicable to each. This has become especially true in our age of 'cooperative federalism,' where the Federal and State Governments are waging a united front against many types of criminal activity." 378 U. S., at 55-56 (citation and internal quotation marks omitted).
Since in this case there is no analog of Malloy, imposing the Fifth Amendment beyond the National Government, there is no premise in Murphy for appealing to "cooperative internationalism" by analogy to "cooperative federalism." 16 Any analogy must, instead, be to the pre-Murphy era when the States were not bound by the privilege. Then, testimony compelled in a federal proceeding was admissible in a state prosecution, despite the fact that shared values and similar criminal statutes of the state and national jurisdictions presumably furnished incentive for overreaching by the Government to facilitate criminal prosecutions in the States.
But even if Murphy were authority for considering "cooperative federalism" and "cooperative internationalism" as reasons supporting expansion of the scope of the privilege,
16 There is indeed nothing comparable to the Fifth Amendment privilege in any supranational prohibition against compelled self-incrimination derived from any source, the privilege being "at best an emerging principle of international law." See Amann, A Whipsaw Cuts Both Ways, 45 UCLA L. Rev. 1201, 1259 (1998) (hereinafter Amann).
In the course of discussing the Eleventh Circuit case raising the same issue as this one, Amann suggests nonetheless that the whipsaw rationale has particular salience on these facts because along with the United States, Lithuania and Israel are signatories to the International Covenant on Civil and Political Rights, Dec. 16, 1966, G. A. Res. 2200, which recognizes something akin to the privilege. See Amann 1233, n. 206. The significance of being bound by the Covenant, however, is limited by its provision that the privilege is derogable and accordingly may be infringed if public emergency necessitates. Id., at 1259, n. 354. In any event, Balsys has made no claim under the Covenant, and its current enforceability in the courts of the signatories is an issue that is not before us.
any extension would depend ultimately on an analysis of the likely costs and benefits of extending the privilege as Balsys requests. If such analysis were dispositive for us, we would conclude that Balsys has not shown that extension of the protection would produce a benefit justifying the rule he seeks.
The Court of Appeals directed careful attention to an evaluation of what would be gained and lost on Balsys's view. It concluded, for example, that few domestic cases would be adversely affected by recognizing the privilege based upon fear of foreign prosecution, 119 F. 3d, at 135-137; 17 that American contempt sanctions for refusal to testify are so lenient in comparison to the likely consequences of foreign prosecution that a witness would probably refuse to testify even if the privilege were unavailable to him, id., at 142 (Block, J., concurring); that by statute and treaty the United States could limit the occasions on which a reasonable fear of foreign prosecution could be shown, as by modifying extradition and deportation standards in cases involving the privilege, id., at 138-139; and that because a witness's refusal to testify may be used as evidence in a civil proceeding, deportation of people in Balsys's position would not necessarily be thwarted by recognizing the privilege as he claims it, id., at 136.
The Court of Appeals accordingly thought the net burden of the expanded privilege too negligible to justify denying its expansion. We remain skeptical, however. While we will not attempt to comment on every element of the Court of Appeals's calculation, two of the points just noted would present difficulty. First, there is a question about the standard that should govern any decision to justify a truly discretionary ruling by making the assumption that it will induce the Government to adopt legislation with international implications or to seek international agreements, in order to
17The assessment was, of course, necessarily based on experience under the same-sovereign view of the privilege.
mitigate the burdens that the ruling would otherwise impose. Because foreign relations are specifically committed by the Constitution to the political branches, Art. II, § 2, cl. 2, we would not make a discretionary judgment premised on inducing them to adopt policies in relation to other nations without squarely confronting the propriety of grounding judicial action on such a premise.
Second, the very assumption that a witness's silence may be used against him in a deportation or extradition proceeding due to its civil nature, 119 F. 3d, at 136 (citing LopezMendoza, 468 U. S., at 1038-1039), raises serious questions about the likely gain from recognizing fear of foreign prosecution. For if a witness claiming the privilege ended up in a foreign jurisdiction that, for whatever reason, recognized no privilege under its criminal law, the recognition of the privilege in the American courts would have gained nothing for the witness. This possibility, of course, presents a sharp contrast with the consequences of recognizing the privilege based on fear of domestic prosecution. If testimony is compelled, Murphy itself illustrates that domestic courts are not even wholly dependent on immunity statutes to see that no use will be made against the witness; the exclusionary principle will guarantee that. See Murphy, 378 U. S., at 79. Whatever the cost to the Government may be, the benefit to the individual is not in doubt in a domestic proceeding.
Since the likely gain to the witness fearing foreign prosecution is thus uncertain, the countervailing uncertainty about the loss of testimony to the United States cannot be dismissed as comparatively unimportant. That some testimony will be lost is highly probable, since the United States will not be able to guarantee immunity if testimony is compelled (absent some sort of cooperative international arrangement that we cannot assume will occur). While the Court of Appeals is doubtless correct that the expected consequences of some foreign prosecutions may be so severe that a witness will refuse to testify no matter what, not
every foreign prosecution may measure up so harshly as against the expectable domestic consequences of contempt for refusing to testify. We therefore must suppose that on Balsys's view some evidence will in fact be lost to the domestic courts, and we are accordingly unable to dismiss the position of the United States in this case, that domestic law enforcement would suffer serious consequences if fear of foreign prosecution were recognized as sufficient to invoke the privilege.
In sum, the most we would feel able to conclude about the net result of the benefits and burdens that would follow from Balsys's view would be a Scotch verdict. If, then, precedent for the traditional view of the scope of the Clause were not dispositive of the issue before us, if extending the scope of the privilege were open to consideration, we still would not find that Balsys had shown that recognizing his claim would be a sound resolution of the competing interests involved.
This is not to say that cooperative conduct between the United States and foreign nations could not develop to a point at which a claim could be made for recognizing fear of foreign prosecution under the Self-Incrimination Clause as traditionally understood. If it could be said that the United States and its allies had enacted substantially similar criminal codes aimed at prosecuting offenses of international character, and if it could be shown that the United States was granting immunity from domestic prosecution for the purpose of obtaining evidence to be delivered to other nations as prosecutors of a crime common to both countries, then an argument could be made that the Fifth Amendment should apply based on fear of foreign prosecution simply because that prosecution was not fairly characterized as distinctly "foreign." The point would be that the prosecution was as much on behalf of the United States as of the prosecuting nation, so that the division of labor between evidence gath-
erer and prosecutor made one nation the agent of the other, rendering fear of foreign prosecution tantamount to fear of a criminal case brought by the Government itself.
Whether such an argument should be sustained may be left at the least for another day, since its premises do not fit this case. It is true that Balsys has shown that the United States has assumed an interest in foreign prosecution, as demonstrated by OS1's mandate 18 and American treaty agreements 19 requiring the Government to give to Lithuania and Israel any evidence provided by Balsys. But this interest does not rise to the level of cooperative prosecution. There is no system of complementary substantive offenses
18 According to Order No. 851-79, reprinted in App. 15-17, the OSI shall "[m]aintain liaison with foreign prosecution, investigation and intelligence offices; [u]se appropriate Government agency resources and personnel for investigations, guidance, information, and analysis; and [d]irect and coordinate the investigation, prosecution, and any other legal actions instituted in these cases with the Immigration and Naturalization Service, the Federal Bureau of Investigation, the United States Attorneys Offices, and other relevant Federal agencies."
19 The United States and Lithuania have entered into an agreement that provides that the two governments "agree to cooperate in prosecution of persons who are alleged to have committed war crimes ... agree to provide mutual legal assistance concerning the prosecution of persons suspected of having committed war crimes ... will assist each other in the location of witnesses believed to possess relevant information about criminal actions ... during World War II, and agree to intermediate and endeavor to make these witnesses available for the purpose of giving testimony in accordance with the laws of the Republic of Lithuania to authorized representatives of the United States Department of Justice." Memorandum of Understanding Between the United States Department of Justice and the Office of the Procurator General of the Republic of Lithuania Concerning Cooperation in the Pursuit of War Criminals, Aug. 3, 1992, reprinted in App. in No. 96-6144 (CA2), pp. 396-397.
The District Court found that though it had not been made aware of a treaty between the United States and Israel requiring disclosure of information related to war crimes, OSI had shared such information in the past and that it would be consistent with OSI's mandate from the Attorney General for OSI to do so again. 918 F. Supp. 588, 596 (EDNY 1996).
at issue here, and the mere support of one nation for the prosecutorial efforts of another does not transform the prosecution of the one into the prosecution of the other. Cf. Bartkus v. Illinois, 359 U. S. 121 , 122-124 (1959) (rejecting double jeopardy claim where federal officials turned over all evidence they had gathered in connection with federal prosecution of defendant for use in subsequent state prosecution of defendant). In this case there is no basis for concluding that the privilege will lose its meaning without a rule precluding compelled testimony when there is a real and substantial risk that such testimony will be used in a criminal prosecution abroad.
Accordingly, the judgment of the Court of Appeals is reversed, and the case is remanded for further proceedings consistent with this opinion.
It is so ordered.
JUSTICE STEVENS, concurring.
While I join the Court's opinion without reservation, I write separately to emphasize these points.
The Clause that protects every person from being "compelled in any criminal case to be a witness against himself" is a part of the broader protection afforded by the Fifth Amendment to the Constitution. That Amendment constrains the power of the Federal Government to deprive any person "of life, liberty, or property, without due process of law," just as the Fourteenth Amendment imposes comparable constraints on the power of the States. The primary office of the Clause at issue in this case is to afford protection to persons whose liberty has been placed in jeopardy in an American tribunal. The Court's holding today will not have any adverse impact on the fairness of American criminal trials.
The fact that the issue in this case has been undecided for such a long period of time suggests that our ruling will have
little, if any, impact on the fairness of trials conducted in other countries. Whether or not that suggestion is accurate, I do not believe our Bill of Rights was intended to have any effect on the conduct of foreign proceedings. If, however, we were to accept respondent's interpretation of the Clause, we would confer power on foreign governments to impair the administration of justice in this country. A law enacted by a foreign power making it a crime for one of its citizens to testify in an American proceeding against another citizen of that country would immunize those citizens from being compelled to testify in our courts. Variants of such a hypothetical law are already in existence. See Societe Nationale Industrielle Aerospatiale v. United States Dist. Court for Southern Dist. of Iowa, 482 U. S. 522 , 526, n. 6 (1987); see also id., at 544-545, n. 29. Of course, the Court might craft exceptions for such foreign criminal laws, but it seems far wiser to adhere to a clear limitation on the coverage of the Fifth Amendment, including its privilege against self-incrimination. That Amendment prescribes rules of conduct that must attend any deprivation of life, liberty, or property in our Nation's courts.
JUSTICE GINSBURG, dissenting.
The privilege against self-incrimination, "closely linked historically with the abolition of torture," is properly regarded as a "landmar[k] in man's struggle to make himself civilized." E. Griswold, The Fifth Amendment Today 7 (1955); see id., at 8 (Fifth Amendment expresses "one of the fundamental decencies in the relation we have developed between government and man"). In my view, the Fifth Amendment privilege against self-incrimination prescribes a rule of conduct generally to be followed by our Nation's officialdom. It counsels officers of the United States (and of any State of the United States) against extracting testimony when the person examined reasonably fears that his words would be used against him in a later criminal prosecution.
As a restraint on compelling a person to bear witness against himself, the Amendment ordinarily should command the respect of United States interrogators, whether the prosecution reasonably feared by the examinee is domestic or foreign. Cf. DKT Memorial Fund Ltd. v. Agency for International Development, 887 F.2d 275 , 307-308 (CADC 1989) (R. B. Ginsburg, J., concurring in part and dissenting in part) ("just as our flag carries its message ... both at home and abroad, so does our Constitution and the values it expresses") (citation and internal quotation marks omitted); United States v. Tiede, 86 F. R. D. 227 (u. S. Court for Berlin 1979) (foreign national accused of hijacking Polish aircraft abroad was tried under German substantive law in Berlin in a court created by United States; U. S. court held foreign national entitled to jury trial as a matter of constitutional right). On this understanding of the "fundamental decenc[y]" the Fifth Amendment embodies, "its expression of our view of civilized governmental conduct," Griswold, supra, at 8, 9, I join JUSTICE BREYER'S dissenting opinion.
JUSTICE BREYER, with whom JUSTICE GINSBURG joins, dissenting.
Were Aloyzas Balsys to face even a theoretical possibility that his testimony could lead a State to prosecute him for murder, the Fifth Amendment would prohibit the Federal Government from compelling that testimony. The Court concludes, however, that the Fifth Amendment does not prohibit compulsion here because Balsys faces a real and substantial danger of prosecution not, say, by California, but by a foreign nation. The Fifth Amendment, however, provides that "[n]o person ... shall be compelled in any criminal case to be a witness against himself." U. S. Const., Amdt. 5 (emphasis added). This Court has not read the words "any criminal case" to limit application of the Clause to only federal criminal cases. See Murphy v. Waterfront Comm'n of N. Y. Harbor, 378 U. S. 52 (1964). That precedent, as well
as the basic principles underlying the privilege, convince me that the Fifth Amendment's privilege against selfincrimination should encompass not only feared domestic prosecutions, but also feared foreign prosecutions where the danger of an actual foreign prosecution is substantial.
I begin with a point that focuses upon precedent setting forth the current understanding of the scope of the word "any," and that reveals the basic difference between the majority's view of the privilege and the view this Court has previously taken and should continue to take. The majority focuses upon one case, Murphy v. Waterfront Comm'n of N. Y. Harbor, supra, which itself discusses much historically relevant precedent. And the majority's focus upon that one case is appropriate.
Murphy holds that "the constitutional privilege against self-incrimination protects ... a federal witness against in-crimination under state ... law." Id., at 77-78. As I read Murphy, the Court thought this conclusion flowed naturally from its basic understanding of the scope of the Fifth Amendment privilege. On that understanding, the privilege prohibits federal courts (and state courts through the Fourteenth Amendment) from compelling a witness to furnish testimonial evidence that may be used to prove his guilt if that witness may reasonably fear criminal prosecution. See id., at 60-63 (discussing the English cases, King of Two Sicilies v. Willcox, 1 Sim. (N. S.) 301, 61 Eng. Rep. 116 (Ch. 1851), and United States of America v. McRae, 3 L. R. Ch. 79 (1867), as ones that, if rightly understood, embody that proposition of law).
The privilege, understood in this way, requires the abolition of any "same sovereign" rule. It is often reasonable for a federal witness to fear state prosecution, and vice versa. Indeed, where testimony may incriminate and immunity has not been granted, it is so reasonable that one can say, as a
matter of law, that the privilege applies, across jurisdictions, to the entire class of cases involving federal witnesses who fear state prosecutions and also to the entire class of cases involving state witnesses who fear federal prosecutions. See Murphy, supra, at 77-78. Thus, the Fifth Amendment (or the Fourteenth Amendment) automatically prohibits compelled testimony in any such cross-jurisdictional circumstance.
If I am right about how Murphy should be understood, then that case directs the application of the privilege in this one. That is because the only difference between Murphy and this case is that one cannot say, as a matter of law, that every threat of a foreign prosecution is a reasonable threat. But where there is such a reasonable threat-where the threat is "real and substantial," Zicarelli v. New Jersey Comm'n of Investigation, 406 U. S. 472 , 478 (1972)-the privilege, as Murphy understands it, would apply.
The majority says that one can read Murphy as embodying a very different rationale, a rationale that turns upon considerations of federalism-the need to consider "state and federal jurisdictions ... as one" for purposes of applying the privilege. Ante, at 683. It reads Murphy as a case that sees at the heart of the Clause
"the principle that the courts of a government from which a witness may reasonably fear prosecution may not in fairness compel the witness to furnish testimonial evidence that may be used to prove his guilt." Ante, at 683 (emphasis added).
I have underscored the key words "from which." It is these words that tie the Clause to prosecutions by the same sovereign.
But what is the evidence that Murphy put any legal weight at all upon those underscored words? What reason
has the majority to believe that Murphy subscribes to, or depends in any way upon, this phrasing of the privilege's "principle" rather than upon the critically different "principle" I suggested above, i. e., the principle that "courts may not in fairness compel a witness who reasonably fears prosecution to furnish testimony that may be used to prove his guilt?"
The majority points to two relevant Murphy statements.
In the first, Murphy said that Malloy v. Hogan, 378 U. S. 1 (1964), which incorporated the Fifth Amendment privilege as part of the Fourteenth Amendment's Due Process Clause, "necessitates a reconsideration" of United States v. Murdock, 284 U. S. 141 (1931), which had held that the Fifth Amendment protected an individual only from prosecutions by the Federal Government. Murphy, 378 U. S., at 57. In the second, Murphy mentioned, as one of many items of support for its analysis, that most Fifth Amendment policies are defeated
"when a witness 'can be whipsawed into incriminating himself under both state and federal law even though' the constitutional privilege against self-incrimination is applicable to each." Id., at 55 (quoting Knapp v. Schweitzer, 357 U. S. 371 , 385 (1958) (Black, J., dissenting) ).
Since the first statement mentions only a reason for reconsidering Murdock, since the second offers support on either analysis, and since neither refers to any "alternative rational[e]" for decision, ante, at 680, the majority's evidence for its reinterpretation of Murphy seems rather skimpy.
Now consider the reasons for believing that Murphy rests upon a different rationale-a rationale that, by focusing upon the basic nature and history of the underlying right, rejects Murdock's "same sovereign" rule. First, Murphy holds that the "constitutional privilege" itself, not that privilege together with principles of federalism, "protects ... a federal
witness against incrimination under state ... law." Murphy, supra, at 78. Second, it says explicitly that it "reject[s]" the Murdock rule, not because of considerations of federalism arising out of Malloy, but because it is "unsupported by history or policy" and represents a "deviation" from a "correct ... construction" of the privilege in light of its "history, policies and purposes." Murphy, supra, at 77. Third, about half of the opinion consists of an effort to demonstrate that the privilege, as understood by the English courts and by American courts prior to Murdock, protected individuals from compelled testimony in the face of a realistic threat of prosecution by any sovereign, not simply by the same sovereign that compelled the testimony. See Murphy, 378 U. S., at 58-70. Fourth, the rest of the Court's analysis consists of a discussion of the purposes of the privilege, which, in the Court's view, lead to a similar conclusion. See id., at 55-56. Fifth, the Court explicitly rejects the analysis of commentators who argued for a "same sovereign" rule on the ground that their understanding of the privilege's purposes was incomplete. See id., at 56-57, n. 5 (rejecting 8 J. Wigmore, Evidence §2258, p. 345 (McNaughton rev. 1961)). Sixth, the Court nowhere describes its rationale in "silver platter" or similar terms that could lead one to conclude that its rule is prophylactic, enforcement based, or rests upon any rationale other than that the privilege is not limited to protection against prosecution by the same jurisdiction that compels the testimony. Cf. 378 U. S., at 80-81 (Harlan, J., concurring in judgment).
Consequently, I believe one must read Murphy as standing for the proposition that the privilege includes protection against being compelled to testify by the Federal Government where that testimony might be used in a criminal prosecution conducted by another sovereign. And the question the Court must consequently face is whether we should reject the rationale of that case when we answer the ques-
tion presented here. In other words, we must ask not, "what did Murphy hold," but "was Murphy right?"
Since Murphy is prevailing law, the majority bears the burden of showing that Murphy is wrong; and the majority says that Murphy's reasoning is "fatally flawed" and legally "unsound." Ante, at 687-688. But it is not. Murphy's reasoning finds in Malloy's holding (that the privilege binds the States) a need to reexamine the "same sovereign" rule, first set forth in the earlier case of Murdock. Without reexamination, Murdock's rule would have permitted State and Federal Governments each to have compelled testimony for use by the other. Murphy's reasoning then finds the "same sovereign" rule unsound as a matter of history and of the basic purposes of the privilege.
Murphy's use of legal history is traditional. It notes that Murdock rested its own conclusion upon earlier English and American cases. It reads the language of those cases in light of the reasons that underlie it. It says that, so read, those cases did not stand for a "same sovereign" rule, but suggested the contrary. And it concludes that Murdock's legal pedigree is suspicious or illegitimate. In a word, Murphy examines Murdock's historical pedigree very much the way that the majority today analyzes that of Murphy. The difference, however, is that Murphy makes a better case for overturning its predecessor than does the majority.
I can reiterate the essence of Murphy's analysis, amending it to fit the present case, roughly as follows:
1. Murdock thought that English law embodied a "same sovereign" rule, but it did not. Two early English cases, one decided in 1749 and the other in 1750, held that the privilege applied even though the feared prosecution was, in the one case, in Calcutta, and in the other, by ecclesiastical authorities. East India Co. v.
Campbell, 1 Yes. sen. 246, 27 Eng. Rep. 1010 (Ex. 1749); Brownsword v. Edwards, 2 Yes. sen. 243, 28 Eng. Rep. 157 (Ch. 1750). Those cases said nothing about whether or not the law of Calcutta, church law, and English law all emanate from a single sovereign. But Murdock had cited a famous later English case, King of Two Sicilies v. Willcox, 1 Sim. (N. S.) 301, 61 Eng. Rep. 116 (Ch. 1851), as standing for the "same sovereign" principle.
It is true that one of the English judges in that case, Lord Cranworth, said that the privilege involves only "matters [made] penal by [English] ... law." Id., at 329, 61 Eng. Rep., at 128. But Lord Cranworth immediately qualified that conclusion by restating the conclusion in terms of its rationale, namely, that the privilege applies "to matters as to which, if disclosed, the Judge would be able to say, as matter of law, whether it could or could not entail penal consequences." Ibid. And, 16 years later, the English courts sustained a claim of privilege involving a threatened forfeiture in America. United States of America v. McRae, 3 L. R. Ch. 79 (1867). In doing so, the McRae court said both that Lord Cranworth's statement in King of the Two Sicilies "la[id] down ... a proposition" that was "broad[er]" than necessary to "support the judgment," and that the true reason the privilege had not applied in the earlier case was because the judge did not "know ... with certainty ... the [foreign law, hence] whether the acts ... had rendered [the defendants] amenable to punishment" and "it was doubtful whether the Defendants would ever be within the reach of a prosecution, and their being so depended on their voluntary return to [Sicily]." United States of America v. McRae, supra, at 85,87.
Thus, the true English rule as of the time of Murdock, insofar as any of these cases reveal that rule, was not a "same sovereign" rule, but a rule that the privilege did not apply to prosecutions by another sovereign where
the danger of any such prosecution was speculative or insubstantial. Cf. Queen v. Boyes, 1 B. & S. 311, 330, 121 Eng. Rep. 730, 738 (Q. B. 1861) ("[T]he danger to be apprehended must be real and appreciable ... not a danger of an imaginary and unsubstantial character").
Where is Murphy's error?
2. Murdock thought that earlier American cases required a "same sovereign" rule, but they did not. To the contrary: Chief Justice Marshall, in United States v. Saline Bank of Va., 1 Pet. 100 (1828), wrote that "a party is not bound to make any discovery which would expose him to penalties." Id., at 104. Justice Holmes later cited this case as authority for the proposition that the Fifth Amendment privilege "exonerated" a federal witness "from [making] disclosures which would have exposed him to the penalties of the state law." BaUmann v. Fagin, 200 U. S. 186 , 195 (1906). Lower federal courts, consistent with the English rule, had held that a witness could refuse to answer questions based on the danger of incrimination in another jurisdiction. See, e. g., In re Hess, 134 F.1d 9 , 112 (ED Pa. 1905); In re Graham, 10 F. Cas. 913, 914 (No. 5,659) (SDNY 1876). True, the Court had written in dicta that "[w]e think the legal immunity is in regard to a prosecution in the same jurisdiction, and when that is fully given it is enough." Jack v. Kansas, 199 U. S. 372 , 382 (1905). But that unexplained dicta, which a later case linked to a (misunderstood) English rule, see Hale v. Henkel, 201 U. S. 43 , 68-69 (1906), provides an insufficient historical basis for Murdock's summary conclusion, particularly since the Court, immediately prior to Murdock, had indicated that the question remained open. See United States ex rel. Vajtauer v. Commissioner of Immigration, 273 U. S. 103 (1927) (reserving question; citing Saline Bank and BaUmann v. Fagin).
Again, where is Murphy's error?
Stated in this minimal way, Murphy's historical analysis is difficult to attack. One can, of course, always point to special features of a case and thereby distinguish it. In respect to the mid-18th-century English cases, one can point out that Calcutta and the church may not have been completely separate "sovereigns." Ante, at 685. And Saline Bank might have involved application by the federal court of a state law that, without the help of the Fifth Amendment, protected a party from self-incrimination. But see Saline Bank, supra, at 103 (citing Virginia privilege statute which, by its terms, applied to suit by the state "Attorney General" in the state "Superior Court of Chancery for the district of Richmond" for recovery of a bank's capital stock "in behalf of the Commonwealth"). But this kind of criticism is beside the point. The English judges made no point of the former. See ante, at 685 (statements about the privilege in these cases were "unqualified"). It does not denigrate their learning to suggest that they did not articulate the precise sovereigntyrelated status of ecclesiastical courts or of Calcutta's criminal law in 1749. Nor did Justice Holmes make any point of the latter. See BaUmann v. Fagin, supra, at 195. As for the suggestion that it is illegitimate to consider the later English authorities in construing the privilege, see ante, at 687, one would think that, on this view, Murdock is at least as vulnerable as Murphy.
Most importantly, neither the majority today, nor the authorities it cites, see ante, at 688-689, n. 11, shows that the key historical points upon which Murphy relied are clearly wrong. At worst, Murphy represents one possible reading of a history that is itself unclear. Murphy's main criticisms of Murdock are reasonable ones. Its reading of earlier cases, insofar as they were relevant to its criticism of Murdock, was plausible then, see Grant, Federalism and SelfIncrimination, 4 UCLA L. Rev. 549, 562 (1957) (Murdock "illustrates the danger of copying one's precedents directly
from the brief of counsel"); and it is plausible now. That minimalist conclusion is sufficient for present purposes. Even if Murdock's 3-sentence, and Murphy's 20-page, historical analyses were equally plausible, we would need something more to abandon Murphy, for it is the most recent, and thereby governing, precedent.
Nor can I find any other reason for rejecting Murphy and, thereby, resurrecting Murdock. The Fifth Amendment's language permits Murphy's construction, for it says "any criminal case." The history of the Amendment's enactment simply does not answer the question about whether or not it applied where there is a substantial danger of prosecution in another jurisdiction. See United States v. Gecas, 120 F.3d 1419 , 1435 (CAll 1997) (en bane) (Fifth Amendment privilege "has virtually no legislative history"); Moglen, Taking the Fifth: Reconsidering the Origins of the Constitutional Privilege Against Self-Incrimination, 92 Mich. L. Rev. 1086, 1123 (1994) (Fifth Amendment's legislative history "adds little to our understanding of the history of the privilege"). It is possible that the language, "in any criminal case," was aimed at limiting protection to compelled testimony against penal interests, a reading consistent with the Court's contemporary understanding of the Clause. See, e. g., United States v. Ward, 448 U. S. 242 , 248-255 (1980) (rejecting claim to privilege based on fear of civil penalty, in part, because Clause "is expressly limited to 'any criminal case' "); 5 The Founders' Constitution 262 (P. Kurland & R. Lerner eds. 1987) (indicating that phrase "in any criminal case" was proposed by Representative Lawrence to ensure that the Clause was not "in some degree contrary to laws passed"). And it is also possible that the language was intended to limit the proceedings in which the privilege could be claimed to criminal cases, which understanding the Court rejected long ago. See McCarthy v. Arndstein, 266 U. S. 34 , 40 (1924) (The privilege "applies alike to civil and criminal proceedings, wherever the answer might tend to subject to criminal responsi-
bility him who gives it"). Neither of these readings is any more speculative, as a textual or historical matter, than reading the Clause as the majority does, against its text, to restrict the universe of feared prosecutions upon which basis the privilege may be asserted.
What is more, there is no suggestion that Murphy's rule, applied to state and federal prosecutions, "has proven to be intolerable simply in defying practical workability." Planned Parenthood of Southeastern Pa. v. Casey, 505 U. S. 833 , 854 (1992) (citing Swift & Co. v. Wickham, 382 U. S. 111 , 116 (1965)). Nor have the facts, or related principles of law, subsequently changed so much "as to have robbed the old rule of significant application or justification." 505 U. S., at 855 (citing Patterson v. McLean Credit Union, 491 U. S. 164 , 173-174 (1989), and Burnet v. Coronado Oil & Gas Co., 285 U. S. 393, 412 (1932) (Brandeis, J., dissenting)). Indeed, it was the Murdock rule's legitimacy that, prior to Murphy, consistently divided the Court. See, e. g., Irvine v. California, 347 U. S. 128 , 139-142 (1954) (Black, J., joined by Douglas, J., dissenting) ("I cannot agree that the [Fifth] Amendment's guarantee against self-incrimination testimony can be spirited away by the ingenious contrivance of using federally extorted confessions to convict of state crimes and vice versa"); Feldman v. United States, 322 U. S. 487 , 494-503 (1944) (Black, J., joined by Douglas and Rutledge, JJ., dissenting).
The conclusion that I draw is that the rationale established through Murphy's precedent governs. That rationale interprets the privilege as applicable at the least where a person faces a substantial threat of prosecution in another jurisdiction. And that reading of the privilege favors Balsys here.
Precedent aside, I still disagree with the Court's conclusion. As Murphy said, and as the Second Circuit reiterated, the Fifth Amendment reflects not one, but several different
purposes. 378 U. S., at 55; 119 F.3d 122 , 129 (1997). And whatever the disagreement about the relative weight to be given each of those purposes or their historical origins, I believe that these purposes argue in favor of the Second Circuit's interpretation. Namely, an interpretation that finds the Fifth Amendment privilege applicable where the threat of a foreign prosecution is "real and substantial," as it is here. See United States of America v. McRae, 3 L. R. Ch., at 85-87 (distinguishing King of the Two Sicilies v. Willcox, 1 Sim. (N. S.) 301, 61 Eng. Rep. 116 (Ch. 1851), on this ground); cf. Queen v. Boyes, 1 B. & S., at 330, 121 Eng. Rep., at 738.
This Court has often found, for example, that the privilege recognizes the unseemliness, the insult to human dignity, created when a person must convict himself out of his own mouth. "At its core, the privilege reflects our fierce 'unwillingness to subject those suspected of crime to the cruel [choice] of self-accusation, perjury or contempt.'" Pennsylvania v. Muniz, 496 U. S. 582 , 596 (1990) (quoting Doe v. United States, 487 U. S. 201 , 212 (1988)); South Dakota v. Neville, 459 U. S. 553 , 563 (1983). The privilege can reflect this value, and help protect against this indignity, even if other considerations produce only partial protection-protection that can be overcome by other needs. Cf. MacNair, Early Development of the Privilege Against SelfIncrimination, 10 Oxford J. Legal Studies 66, 70 (1990) (early ecclesiastical procedure recognized privilege until an accusation was made that person had committed an offense); ante, at 692 (observing that the "protection of personal testimonial inviolability" is not a "reliable guid[e]" to the "actual scope of protection under the Clause"). And that value is no less at stake where a foreign, but not a domestic, prosecution is at issue.
This Court has also said that the privilege serves to protect personal privacy, by discouraging prosecution for crimes
of thought. See Muniz, supra, at 595-596 (describing English Star Chamber "wherein suspects were forced to choose between revealing incriminating private thoughts and forsaking their oath by committing perjury"); United States v. Nobles, 422 U. S. 225 , 233 (1975) ("The Fifth Amendment privilege ... protects 'a private inner sanctum of individual feeling and thought and proscribes state intrusion to extract self-condemnation'" (quoting Couch v. United States, 409 U. S. 322, 327 (1973))). Indeed, some have argued that the Puritans championed the privilege because, had the 17thcentury state questioned them about their beliefs, they would have had to answer truthfully and thus suffer condemnation. See L. Levy, Origins of the Fifth Amendment 134 (1968) ("If [a Puritan] took the oath and lied, he committed the unpardonable and cardinal sin of perjury which was simply not an option for a religious man"). This consideration may prove less important today domestically, for the First Amendment protects against the prosecution of thought crime. But that fact also provides no reason for denying protection where the prosecution is foreign.
The Court has said that the privilege reflects, too, "our fear that self-incriminating statements will be elicited by inhumane treatment and abuses." Murphy, 378 U. S., at 55. This concern with governmental "overreaching" would appear implicated as much when the foreseen prosecution is by another country as when it is by another domestic jurisdiction. Indeed, the analogy to Murphy's observation about "cooperative federalism," in which State and Federal Governments wage "a united front against many types of criminal activity," id., at 56, is a powerful one. That is because, in the 30 years since Murphy, the United States has dramatically increased its level of cooperation with foreign governments to combat crime. See generally E. N adelman, Cops Across Borders: The Internationalization of U. S. Criminal Law Enforcement (1993); Bassiouni, Policy Considerations on Inter-State Cooperation in Criminal Matters, 4 Pace Y. B.
Int'l L. 123 (1992); Zagaris, International Criminal and Enforcement Cooperation in the Americas in the Wake of Integration, 3 Sw. J. L. & Trade Am. 1 (1996). The United States has entered into some 20 "mutual legal assistance treaties" through which it may develop and share evidence with foreign governments in order to facilitate criminal prosecutions abroad, see New MLAT Treaties Increase DOJ's Reach, 4 No.7 DOJ Alert 7 (Apr. 18, 1994) (listing and discussing treaties); it has signed more than 50 new extradition agreements, see 18 U. S. C. § 3181 (1994 ed., Supp. II) (listing extradition treaties ratified since 1960); Nadelman, Cops Across Borders, at 489-502 (same); it has increased by an order of magnitude the number of law enforcement offices and personnel located abroad, see id., at 479-486 (cataloging growth in foreign-based law enforcement personnel since 1965); and it has established a special office" 'for the purpose of centralizing and giving greater emphasis and visibility to [the Justice Department's] prosecutorial service functions in the international arena,'" which has led to a "dramatic increase in the number of extraditions" and an "even greater growth in the numbers of requests for evidence in criminal cases" since the 1970's, id., at 402 (discussing DOJ's Office of International Affairs (alterations omitted)).
Indeed, the United States has a significant stake in the foreign prosecution at issue here. Congress has passed a deportation law targeted at suspected Nazi war criminals. See 8 U. S. C. § 1182(a)(3)(E). The Justice Department has established an agency whose mandate includes the assistance of foreign governments in the prosecution of those deported. See App. 15-17 (Order No. 851-79, establishing DOJ's Office of Special Investigations). And the United States has agreed with Lithuania (where Balsys may stand trial) "to cooperate in prosecution of persons who are alleged to have committed war crimes ... [and] to provide ... legal assistance concerning [such] prosecution[s]." Memorandum of Understanding Between United States Department of Jus-
tice and Office of Procurator General of the Republic of Lithuania Concerning Cooperation in the Pursuit of War Criminals, Aug. 3, 1992, App. in No. 96-6144 (CA2), p. 395. As the Second Circuit reasoned, since the Federal Government now has a stake in many foreign prosecutions akin to its stake in state prosecutions, a stake illustrated by this case, the privilege's purpose of preventing governmental overreaching is served by recognizing the privilege in the former class of cases, just as it is served in the cases of "cooperative federalism" identified by Murphy. Indeed, experience suggests that the possibility of governmental abuses in cases like this one-where the United States has an admittedly keen interest in the later, foreign prosecution-is not totally speculative. See, e. g., Demjanjuk v. Petrovsky, 10 F.3d 338 (CA61993).
An additional purpose served by the privilege is "our preference for an accusatorial rather than an inquisitorial system of criminal justice." Murphy, supra, at 55. Even if this systemic value speaks to "domestic arrangements" only, ante, at 690, the investigation of crime is as much a part of our "system" of criminal justice as is any later criminal prosecution. Reflecting this fact, the Court has said that the Fifth Amendment affords individuals protection during the investigation, as well as the trial, of a crime. See Mi randa v. Arizona, 384 U. S. 436 (1966). And the importance we place in our system of criminal investigation, and the distaste we have for its alternatives, would stand diminished if an accused were denied the Fifth Amendment's protections because the criminal case against him, though built in this country by our Government, was ultimately to be prosecuted in another. This is true regardless of whether the "Bill of Rights was intended to have any effect on the conduct of foreign proceedings." Ante, at 701 (STEVENS, J., concurring). The Fifth Amendment undeniably "prescribes a rule of conduct generally to be followed by our Nation's official-
dom," ibid. (GINSBURG, J., dissenting), and it is that conduct, not a foreign proceeding, that is at issue here.
If the policies and purposes that this Court has said underlie the Fifth Amendment-respect for individual dignity and privacy, prevention of governmental overreaching, preservation of an accusatorial system of criminal justice-would all be well served by applying the privilege when a witness legitimately fears foreign prosecution, then what reason could there be for reinterpreting the privilege so as not to recognize it here?
Two reasons have been suggested: First, one might see a government's compulsion of testimony followed by its own use of that testimony in a criminal prosecution as somewhat more unfair than compulsion by one government and use by another. And one might also find the States and the Federal Government so closely interconnected that the unfairness is further diminished where the prosecuting sovereign is a foreign country.
But this factor, in my view, cannot be determinative. For one thing, this issue of fairness is a matter of degree, not kind. For another, changes in transportation and communication have made relationships among nations ever closer, to the point where cooperation among international prosecutors and police forces may be as great today as among the States (or between the States and the Federal Government) a half century ago. See supra, at 714-715 (discussing rise in international cooperation). Finally, this Court's cases suggest that the remaining considerations-particularly the inherent indignity and cruelty to the individual in compelling selfincrimination-bulk larger in terms of the basic values that the Fifth Amendment reflects than does this single, partial, fairness consideration. See supra, at 712-713 (citing cases). I cannot agree that this particular feature-the fact that
prosecution by a different sovereign seems not quite as unfair as prosecution by the same sovereign-could warrant denying the privilege's application.
The second consideration is practical. The majority, as well as the Government, fear that application of the privilege might unreasonably interfere with the work of law enforcement. See ante, at 697-698; Brief for United States 30-36. But in my view, that fear is overstated. After all, "foreign application" of the privilege would matter only in a case where an individual could not be prosecuted domestically but the threat of foreign prosecution is substantial. Cf. Zicarelli, 406 U. S., at 478-481 (declining to reach privilege claim because witness did not face "real danger" of foreign prosecution). The Second Circuit points out that there have only been a handful of such cases. 119 F. 3d, at 135 (finding only six cases in the 25 years since Zicarelli). That is because relatively few witnesses face deportation or extradition, and a witness who will not "'be forced to enter a country disposed to prosecute him,'" 119 F. 3d, at 135 (quoting United States v. Gecas, 50 F.3d 1549 , 1560 (CA111995), cannot make the showing of "real and substantial" fear that Zicarelli would require.
Moreover, even where a substantial likelihood of foreign prosecution can be shown, the Government would only be deprived of testimony that relates to the foreign crime; the witness would not be entitled to claim a general silence. See Hoffman v. United States, 341 U. S. 479 , 486 (1951) (witness may only refuse to answer questions that might "in themselves support a conviction" or "furnish a link in the chain of evidence" for such crime). And nothing would prevent the Government, in a civil proceeding, from arguing that an adverse inference should be drawn from the witnesses' silence on particular questions, see Baxter v. Palmigiano, 425 U. S. 308 , 318 (1976), or from supporting that inference with evidence from other, nonprivileged sources. Thus, without any adjustment in practice, it would seem that
the Government would lose little information, and even fewer cases, were the privilege recognized here.
In those rare instances where the need for testimony was sufficiently great, a grant of de facto "immunity" remains a possibility. The Government need only take steps sufficient to make the threat of foreign prosecution insubstantial. Thus, a promise by the United States that deportation will not take place, or that deportation to a different country will ensue, would seem sufficient. A further promise by the foreign nation that prosecution will not take place, or will not make use of the elicited testimony, will obviate the need even for such a deportation promise. And were a foreign sovereign to later seek extradition of the witness, the Government, under existing law, might retain the discretion to decline such a request. See 18 U. S. C. § 3186 ("Secretary of State may order" extraditable person "delivered to ... foreign government"); § 3196 (giving Secretary of State discretion whether to extradite United States citizens provided treaty does not obligate her to do so).
I do not want to minimize the potential difficulties inherent in providing this kind of "immunity." It might require a change in domestic law, or in a given case, an adjustment in an understanding reached with a foreign government. In unusual circumstances, as JUSTICE STEVENS recognizes, see ante, at 701, it might require adjusting the legal rules that express the privilege in order to prevent a foreign government's efforts to stop its citizens from testifying in American courts. But I do not see these difficulties as creating overwhelming obstacles to the legitimate application of the privilege in instances such as the one present here. Nor do I see these difficulties as significantly greater than those that inhere in the ordinary grant of immunity, which also requires legislation, and which also can create friction among competing jurisdictions. At worst, granting de facto "immunity" in this type of case would mean more potentially deportable criminal aliens will remain in the United States, just as to-
day's immunity means more potentially imprisonable citizens remain at liberty. This is a price that the Amendment extracts where government wishes to compel incriminating testimony; and it is difficult to see why that price should not be paid where there is a real threat of prosecution, but it is foreign.
In sum, I see no reason why the Court should resurrect the pale shadow of Murdock's "same sovereign" rule, a rule that Murphy demonstrated was without strong historical foundation and that would serve no more valid a purpose in to day's world than it did during Murphy's time. Murphy supports recognizing the privilege where there is a real and substantial threat of prosecution by a foreign government. Balsys is among the few to have satisfied this threshold. The basic values that this Court has said underlie the Fifth Amendment's protections are each diminished if the privilege may not be claimed here. And surmountable practical concerns should not stand in the way of constitutional principle.
For these and related reasons elaborated by the Second Circuit, I respectfully dissent.