No. 17-646

Argued: December 6, 2018Decided: June 17, 2019

Petitioner Gamble pleaded guilty to a charge of violating Alabama’s felon-in-possession-of-a-firearm statute. Federal prosecutors then indicted him for the same instance of possession under federal law. Gamble moved to dismiss, arguing that the federal indictment was for “the same offence” as the one at issue in his state conviction, thus exposing him to double jeopardy under the Fifth Amendment. The District Court denied this motion, invoking the dual-sovereignty doctrine, according to which two offenses “are not the ‘same offence’ ” for double jeopardy purposes if “prosecuted by different sovereigns,” Heath v. Alabama, 474 U. S. 82, 92. Gamble pleaded guilty to the federal offense but appealed on double jeopardy grounds. The Eleventh Circuit affirmed.

Held: This Court declines to overturn the longstanding dual-sovereignty doctrine. Pp. 3-31.

(a) The dual-sovereignty doctrine is not an exception to the double jeopardy right but follows from the Fifth Amendment’s text. The Double Jeopardy Clause protects individuals from being “twice put in jeopardy” “for the same offence.” As originally understood, an “offence” is defined by a law, and each law is defined by a sovereign. Thus, where there are two sovereigns, there are two laws and two “offences.” Gamble attempts to show from the Clause’s drafting history that Congress must have intended to bar successive prosecutions regardless of the sovereign bringing the charge. But even if conjectures about subjective goals were allowed to inform this Court’s reading of the text, the Government’s contrary arguments on that score would prevail. Pp. 3-5.

(b) This Court’s cases reflect the sovereign-specific reading of the phrase “same offence.” Three antebellum cases–Fox v. Ohio, 5 How. 410; United States v. Marigold, 9 How. 560; and Moore v. Illinois, 14 How. 13–laid the foundation that a crime against two sovereigns constitutes two offenses because each sovereign has an interest to vindicate. Seventy years later, that foundation was cemented in United States v. Lanza, 260 U. S. 377, which upheld a federal prosecution that followed one by a State. This Court applied that precedent for decades until 1959, when it refused two requests to reverse course, see Bartkus v. Illinois, 359 U. S. 121; Abbate v. United States, 359 U. S. 187, and it has reinforced that precedent over the following six decades, see, e.g.Puerto Rico v. Sanchez Valle, 579 U. S. ___. Pp. 5-10.

(c) Gamble claims that this Court’s precedent contradicts the common-law rights that the Double Jeopardy Clause was originally understood to engraft onto the Constitution, pointing to English and American cases and treatises. A departure from precedent, however, “demands special justification,” Arizona v. Rumsey, 467 U. S. 203, 212, and Gamble’s historical evidence is too feeble to break the chain of precedent linking dozens of cases over 170 years. This Court has previously concluded that the probative value of early English decisions on which Gamble relies was “dubious” due to “confused and inadequate reporting.” Bartkus, 359 U. S., at 128, n. 9. On closer inspection, that assessment has proven accurate; the passing years have not made those early cases any clearer or more valuable. Nor do the treatises cited by Gamble come close to settling the historical question with enough force to meet his particular burden. His position is also not supported by state court cases, which are equivocal at best. Less useful still are the two federal cases cited by Gamble–Houston v. Moore, 5 Wheat. 1, which squares with the dual-sovereignty doctrine, and United States v. Furlong, 5 Wheat. 184, which actually supports it. Pp. 11-28.

(d) Gamble’s attempts to blunt the force of stare decisis here do not succeed. He contends that the recognition of the Double Jeopardy Clause’s incorporation against the States washed away any theoretical foundation for the dual-sovereignty rule. But this rule rests on the fact that only same-sovereign prosecutions can involve the “same offence,” and that is just as true after incorporation as before. Gamble also argues that the proliferation of federal criminal laws has raised the risk of successive prosecutions under state and federal law for the same criminal conduct, thus compounding the harm inflicted by precedent. But this objection obviously assumes that precedent was erroneous from the start, so it is only as strong as the historical arguments found wanting. In any case, eliminating the dual-sovereignty rule would do little to trim the reach of federal criminal law or prevent many successive state and federal prosecutions for the same criminal conduct, see Blockburger v. United States, 284 U. S. 299. Pp. 28-31.

694 Fed. Appx. 750, affirmed.

ALITO, J., delivered the opinion of the Court, in which ROBERTS, C. J., and THOMASBREYERSOTOMAYORKAGAN, and KAVANAUGH, JJ., joined. THOMAS, J., filed a concurring opinion. GINSBURG, J., and GORSUCH, J., filed dissenting opinions.

Opinion of the Court

 587 U. S. ____ (2019)

NOTICE: This opinion is subject to formal revision before publication in the preliminary print of the United States Reports. Readers are requested to notify the Reporter of Decisions, Supreme Court of the United States, Washington, D. C. 20543, of any typographical or other formal errors, in order that corrections may be made before the preliminary print goes to press.

No. 17-646



[June 17, 2019]

JUSTICE ALITO delivered the opinion of the Court.

We consider in this case whether to overrule a longstanding interpretation of the Double Jeopardy Clause of the Fifth Amendment. That Clause provides that no person may be “twice put in jeopardy” “for the same offence.” Our double jeopardy case law is complex, but at its core, the Clause means that those acquitted or convicted of a particular “offence” cannot be tried a second time for the same “offence.” But what does the Clause mean by an “offence”?

We have long held that a crime under one sovereign’s laws is not “the same offence” as a crime under the laws of another sovereign. Under this “dual-sovereignty” doctrine, a State may prosecute a defendant under state law even if the Federal Government has prosecuted him for the same conduct under a federal statute.

Or the reverse may happen, as it did here. Terance Gamble, convicted by Alabama for possessing a firearm as a felon, now faces prosecution by the United States under its own felon-in-possession law. Attacking this second prosecution on double jeopardy grounds, Gamble asks us to overrule the dual-sovereignty doctrine. He contends that it departs from the founding-era understanding of the right enshrined by the Double Jeopardy Clause. But the historical evidence assembled by Gamble is feeble; pointing the other way are the Clause’s text, other historical evidence, and 170 years of precedent. Today we affirm that precedent, and with it the decision below.


     In November 2015, a local police officer in Mobile, Alabama, pulled Gamble over for a damaged headlight. Smelling marijuana, the officer searched Gamble’s car, where he found a loaded 9-mm handgun. Since Gamble had been convicted of second-degree robbery, his possession of the handgun violated an Alabama law providing that no one convicted of “a crime of violence” “shall own a firearm or have one in his or her possession.” Ala. Code §13A-11-72(a) (2015); see §13A-11-70(2) (defining “crime of violence” to include robbery). After Gamble pleaded guilty to this state offense, federal prosecutors indicted him for the same instance of possession under a federal law–one forbidding those convicted of “a crime punishable by imprisonment for a term exceeding one year . . . to ship or transport in interstate or foreign commerce, or possess in or affecting commerce, any firearm or ammunition.” 18 U. S. C. §922(g)(1).

Gamble moved to dismiss on one ground: The federal indictment was for “the same offence” as the one at issue in his state conviction and thus exposed him to double jeopardy. But because this Court has long held that two offenses “are not the ‘same offence’ ” for double jeopardy purposes if “prosecuted by different sovereigns,” Heath v. Alabama, 474 U. S. 82, 92 (1985), the District Court denied Gamble’s motion to dismiss. Gamble then pleaded guilty to the federal offense while preserving his right to challenge the denial of his motion to dismiss on double jeopardy grounds. But on appeal the Eleventh Circuit affirmed, citing the dual-sovereignty doctrine. 694 Fed. Appx. 750 (2017). We granted certiorari to determine whether to overturn that doctrine.1 585 U. S. ___ (2018).


     Gamble contends that the Double Jeopardy Clause must forbid successive prosecutions by different sovereigns because that is what the founding-era common law did. But before turning to that historical claim, see Part III infra, we review the Clause’s text and some of the cases Gamble asks us to overturn.


     We start with the text of the Fifth Amendment. Although the dual-sovereignty rule is often dubbed an “exception” to the double jeopardy right, it is not an exception at all. On the contrary, it follows from the text that defines that right in the first place. “[T]he language of the Clause . . . protects individuals from being twice put in jeopardy ‘for the same offence,’ not for the same conduct or actions,” Grady v. Corbin, 495 U. S. 508, 529 (1990), as Justice Scalia wrote in a soon-vindicated dissent, see United States v. Dixon, 509 U. S. 688 (1993) (overruling Grady). And the term ” ‘[o]ffence’ was commonly understood in 1791 to mean ‘transgression,’ that is, ‘the Violation or Breaking of a Law.’ ” Grady, 495 U. S., at 529 (Scalia, J., dissenting) (quoting Dictionarium Britannicum (Bailey ed. 1730)). See also 2 R. Burn & J. Burn, A New Law Dictionary 167 (1792) (“OFFENCE, is an act committed against law, or omitted where the law requires it”). As originally understood, then, an “offence” is defined by a law, and each law is defined by a sovereign. So where there are two sovereigns, there are two laws, and two “offences.” See Grady, 495 U. S., at 529 (Scalia, J., dissenting) (“If the same conduct violates two (or more) laws, then each offense may be separately prosecuted”); Moore v. Illinois, 14 How. 13, 17 (1852) (“The constitutional provision is not, that no person shall be subject, for the same act, to be twice put in jeopardy of life or limb; but for the same offence, the same violation of law, no person’s life or limb shall be twice put in jeopardy” (emphasis added)).

Faced with this reading, Gamble falls back on an episode from the Double Jeopardy Clause’s drafting history.2 The first Congress, working on an earlier draft that would have banned ” ‘more than one trial or one punishment for the same offence,’ ” voted down a proposal to add ” ‘by any law of the United States.’ ” 1 Annals of Cong. 753 (1789). In rejecting this addition, Gamble surmises, Congress must have intended to bar successive prosecutions regardless of the sovereign bringing the charge.

Even if that inference were justified–something that the Government disputes–it would count for little. The private intent behind a drafter’s rejection of one version of a text is shoddy evidence of the public meaning of an altogether different text. Cf. United States v. Craft, 535 U. S. 274, 287 (2002) (“[F]ailed legislative proposals are a particularly dangerous ground on which to rest an interpretation of a prior statute” (internal quotation marks omitted)).

Besides, if we allowed conjectures about purpose to inform our reading of the text, the Government’s conjecture would prevail. The Government notes that the Declaration of Independence denounced King George III for “protecting [British troops] by a mock Trial, from punishment for any Murders which they should commit on the Inhabitants of these States.” ¶ 17. The Declaration was alluding to “the so-called Murderers’ Act, passed by Parliament after the Boston Massacre,” Amar, Sixth Amendment First Principles, 84 Geo. L. J. 641, 687, n. 181 (1996), a law that allowed British officials indicted for murder in America to be ” ‘tried in England, beyond the control of local juries.’ ” Ibid. (quoting J. Blum et al., The National Experience 95 (3d ed. 1973)). “During the late colonial period, Americans strongly objected to . . . [t]his circumvention of the judgment of the victimized community.” Amar, 84 Geo. L. Rev., at 687, n. 181. Yet on Gamble’s reading, the same Founders who quite literally revolted against the use of acquittals abroad to bar criminal prosecutions here would soon give us an Amendment allowing foreign acquittals to spare domestic criminals. We doubt it.

We see no reason to abandon the sovereign-specific reading of the phrase “same offence,” from which the dual-sovereignty rule immediately follows.


     Our cases reflect the same reading. A close look at them reveals how fidelity to the Double Jeopardy Clause’s text does more than honor the formal difference between two distinct criminal codes. It honors the substantive differences between the interests that two sovereigns can have in punishing the same act.

The question of successive federal and state prosecutions arose in three antebellum cases implying and then spelling out the dual-sovereignty doctrine. The first, Fox v. Ohio, 5 How. 410 (1847), involved an Ohio prosecution for the passing of counterfeit coins. The defendant argued that since Congress can punish counterfeiting, the States must be barred from doing so, or else a person could face two trials for the same offense, contrary to the Fifth Amendment. We rejected the defendant’s premise that under the Double Jeopardy Clause “offences falling within the competency of different authorities to restrain or punish them would not properly be subjected to the consequences which those authorities might ordain and affix to their perpetration.” Id., at 435. Indeed, we observed, the nature of the crime or its effects on “public safety” might well “deman[d]” separate prosecutions. Ibid. Generalizing from this point, we declared in a second case that “the same act might, as to its character and tendencies, and the consequences it involved, constitute an offence against both the State and Federal governments, and might draw to its commission the penalties denounced by either, as appropriate to its character in reference to each.” United States v. Marigold, 9 How. 560, 569 (1850).

A third antebellum case, Moore v. Illinois, 14 How. 13, expanded on this concern for the different interests of separate sovereigns, after tracing it to the text in the manner set forth above. Recalling that the Fifth Amendment prohibits double jeopardy not “for the same ac[t]” but “for the same offence,” and that “[a]n offence, in its legal signification, means the transgression of a law,” id., at 19, we drew the now-familiar inference: A single act “may be an offence or transgression of the laws of” two sovereigns, and hence punishable by both, id., at 20. Then we gave color to this abstract principle–and to the diverse interests it might vindicate–with an example. An assault on a United States marshal, we said, would offend against the Nation and a State: the first by “hindering” the “execution of legal process,” and the second by “breach[ing]” the “peace of the State.” Ibid. That duality of harm explains how “one act” could constitute “two offences, for each of which [the offender] is justly punishable.” Ibid.

This principle comes into still sharper relief when we consider a prosecution in this country for crimes committed abroad. If, as Gamble suggests, only one sovereign may prosecute for a single act, no American court–state or federal–could prosecute conduct already tried in a foreign court. Imagine, for example, that a U. S. national has been murdered in another country. That country could rightfully seek to punish the killer for committing an act of violence within its territory. The foreign country’s interest lies in protecting the peace in that territory rather than protecting the American specifically. But the United States looks at the same conduct and sees an act of violence against one of its nationals, a person under the particular protection of its laws. The murder of a U. S. national is an offense to the United States as much as it is to the country where the murder occurred and to which the victim is a stranger. That is why the killing of an American abroad is a federal offense that can be prosecuted in our courts, see 18 U. S. C. §2332(a)(1), and why customary international law allows this exercise of jurisdiction.

There are other reasons not to offload all prosecutions for crimes involving Americans abroad. We may lack confidence in the competence or honesty of the other country’s legal system. Less cynically, we may think that special protection for U. S. nationals serves key national interests related to security, trade, commerce, or scholarship. Such interests might also give us a stake in punishing crimes committed by U. S. nationals abroad–especially crimes that might do harm to our national security or foreign relations. See, e.g., §2332a(b) (bombings). These examples reinforce the foundation laid in our antebellum cases: that a crime against two sovereigns constitutes two offenses because each sovereign has an interest to vindicate.

We cemented that foundation 70 years after the last of those antebellum cases, in a decision upholding a federal prosecution that followed one by a State. See United States v. Lanza, 260 U. S. 377, 382 (1922) (“[A]n act denounced as a crime by both national and state sovereignties is an offense against the peace and dignity of both and may be punished by each”). And for decades more, we applied our precedent without qualm or quibble. See, e.g.Screws v. United States, 325 U. S. 91 (1945); Jerome v. United States, 318 U. S. 101 (1943); Puerto Rico v. Shell Co. (P. R.), Ltd., 302 U. S. 253 (1937); Westfall v. United States, 274 U. S. 256 (1927); Hebert v. Louisiana, 272 U. S. 312 (1926). When petitioners in 1959 asked us twice to reverse course, we twice refused, finding “[n]o consideration or persuasive reason not presented to the Court in the prior cases” for disturbing our “firmly established” doctrine. Abbate v. United States, 359 U. S. 187, 195; see also Bartkus v. Illinois, 359 U. S. 121. And then we went on enforcing it, adding another six decades of cases to the doctrine’s history. See, e.g.Puerto Rico v. Sánchez Valle, 579 U. S. ___ (2016); Heath v. Alabama, 474 U. S. 82 (1985); United States v. Wheeler, 435 U. S. 313 (1978); Rinaldi v. United States, 434 U. S. 22 (1977) ( per curiam).


     We briefly address two objections to this analysis.

First, the dissents contend that our dual-sovereignty rule errs in treating the Federal and State Governments as two separate sovereigns when in fact sovereignty belongs to the people. See post, at 3 (opinion of GINSBURG, J.); post, at 7 (opinion of GORSUCH, J.). This argument is based on a non sequitur. Yes, our Constitution rests on the principle that the people are sovereign, but that does not mean that they have conferred all the attributes of sovereignty on a single government. Instead, the people, by adopting the Constitution, ” ‘split the atom of sovereignty.’ ” Alden v. Maine, 527 U. S. 706, 751 (1999) (alteration omitted) (internal quotation marks and citation omitted). As we explained last Term:

“When the original States declared their independence, they claimed the powers inherent in sovereignty . . . . The Constitution limited but did not abolish the sovereign powers of the States, which retained ‘a residuary and inviolable sovereignty.’ The Federalist No. 39, p. 245 (C. Rossiter ed. 1961). Thus, both the Federal Government and the States wield sovereign powers, and that is why our system of government is said to be one of ‘dual sovereignty.’ Gregory v. Ashcroft, 501 U. S. 452, 457 (1991).” Murphy v. National Collegiate Athletic Assn., 584 U. S. ___, ___ (2018) (slip op., at 14).

It is true that the Republic is ” ‘ONE WHOLE,’ ” post, at 3 (opinion of GINSBURG, J.) (quoting The Federalist No. 82, p. 493 (C. Rossiter ed. 1961) (A. Hamilton)); accord, post, at 7 (opinion of GORSUCH, J.). But there is a difference between the whole and a single part, and that difference underlies decisions as foundational to our legal system as McCulloch v. Maryland, 4 Wheat. 316 (1819). There, in terms so directly relevant as to seem presciently tailored to answer this very objection, Chief Justice Marshall distinguished precisely between “the people of a State” and “[t]he people of all the States,” id., at 428, 435; between the “sovereignty which the people of a single state possess” and the sovereign powers “conferred by the people of the United States on the government of the Union,” id., at 429-430; and thus between “the action of a part” and “the action of the whole,” id., at 435-436. In short, McCulloch‘s famous holding that a State may not tax the national bank rested on a recognition that the States and the Nation have different “interests” and “right[s].” Id., 431, 436. One strains to imagine a clearer statement of the premises of our dual-sovereignty rule, or a more authoritative source. The United States is a federal republic; it is not, contrary to JUSTICE GORSUCH‘s suggestion, post, at 10-11, a unitary state like the United Kingdom.

Gamble and the dissents lodge a second objection to this line of reasoning. They suggest that because the division of federal and state power was meant to promote liberty, it cannot support a rule that exposes Gamble to a second sentence. See post, at 3-4 (opinion of GINSBURG, J.); post, at 8-9 (opinion of GORSUCH, J.). This argument fundamentally misunderstands the governmental structure established by our Constitution. Our federal system advances individual liberty in many ways. Among other things, it limits the powers of the Federal Government and protects certain basic liberties from infringement. But because the powers of the Federal Government and the States often overlap, allowing both to regulate often results in two layers of regulation. Taxation is an example that comes immediately to mind. It is also not at all uncommon for the Federal Government to permit activities that a State chooses to forbid or heavily restrict–for example, gambling and the sale of alcohol. And a State may choose to legalize an activity that federal law prohibits, such as the sale of marijuana. So while our system of federalism is fundamental to the protection of liberty, it does not always maximize individual liberty at the expense of other interests. And it is thus quite extraordinary to say that the venerable dual-sovereignty doctrine represents a ” ‘desecrat[ion]’ ” of federalism. Post, at 9 (opinion of GORSUCH, J.).


     Gamble claims that our precedent contradicts the common-law rights that the Double Jeopardy Clause was originally understood to engraft onto the Constitution–rights stemming from the “common-law pleas of auterfoits acquit [former acquittal] and auterfoits convict [former conviction].” Grady, 495 U. S., at 530 (Scalia, J., dissenting). These pleas were treated as “reason[s] why the prisoner ought not to answer [an indictment] at all, nor put himself upon his trial for the crime alleged.” 4 W. Blackstone, Commentaries on the Laws of England 335 (1773) (Blackstone). Gamble argues that those who ratified the Fifth Amendment understood these common-law principles (which the Amendment constitutionalized) to bar a domestic prosecution following one by a foreign nation. For support, he appeals to early English and American cases and treatises. We have highlighted one hurdle to Gamble’s reading: the sovereign-specific original meaning of “offence.” But the doctrine of stare decisis is another obstacle.

Stare decisis “promotes the evenhanded, predictable, and consistent development of legal principles, fosters reliance on judicial decisions, and contributes to the actual and perceived integrity of the judicial process.” Payne v. Tennessee, 501 U. S. 808, 827 (1991). Of course, it is also important to be right, especially on constitutional matters, where Congress cannot override our errors by ordinary legislation. But even in constitutional cases, a departure from precedent “demands special justification.” Arizona v. Rumsey, 467 U. S. 203, 212 (1984). This means that something more than “ambiguous historical evidence” is required before we will “flatly overrule a number of major decisions of this Court.” Welch v. Texas Dept. of Highways and Public Transp., 483 U. S. 468, 479 (1987). And the strength of the case for adhering to such decisions grows in proportion to their “antiquity.” Montejo v. Louisiana, 556 U. S. 778, 792 (2009). Here, as noted, Gamble’s historical arguments must overcome numerous “major decisions of this Court” spanning 170 years. In light of these factors, Gamble’s historical evidence must, at a minimum, be better than middling.

And it is not. The English cases are a muddle. Treatises offer spotty support. And early state and federal cases are by turns equivocal and downright harmful to Gamble’s position. All told, this evidence does not establish that those who ratified the Fifth Amendment took it to bar successive prosecutions under different sovereigns’ laws–much less do so with enough force to break a chain of precedent linking dozens of cases over 170 years.


     Gamble’s core claim is that early English cases reflect an established common-law rule barring domestic prosecution following a prosecution for the same act under a different sovereign’s laws. But from the very dawn of the common law in medieval England until the adoption of the Fifth Amendment in 1791, there is not one reported decision barring a prosecution based on a prior trial under foreign law. We repeat: Gamble has not cited and we have not found a single pre-Fifth Amendment case in which a foreign acquittal or conviction barred a second trial in a British or American court. Given this void, Gamble faces a considerable challenge in convincing us that the Fifth Amendment was originally understood to establish such a bar.

Attempting to show that such a bar was available, Gamble points to five early English decisions for which we have case reports. We will examine these in some detail, but we note at the outset that they play only a secondary role for Gamble.

The foundation of his argument is a decision for which we have no case report: the prosecution in England in 1677 of a man named Hutchinson. (We have a report of a decision denying Hutchinson bail but no report of his trial.) As told by Gamble, Hutchinson, having been tried and acquitted in a foreign court for a murder committed abroad, was accused of the same homicide in an English tribunal, but the English court held that the foreign prosecution barred retrial.

Everything for Gamble stems from this one unreported decision. To the extent that the cases he cites provide any support for his argument–and for the most part, they do not–those cases purport to take their cue from the Hutchinson episode; the same is true of the treatises on which Gamble relies.

So what evidence do we have about what actually happened to Hutchinson? The most direct evidence is a report of his application for bail before the Court of King’s Bench. The report spans all of one sentence:

     “On Habeas Corpus it appeared the Defendant was committed to Newgate on suspicion of Murder in Portugal, which by Mr. Attorny being a Fact out of the Kings Dominions, is not triable by Commission, upon 35 H. 8. Cap. 2. §. I. N. 2. but by a Constable and Marshal, and the Court refused to Bail him, & c.” Rex v. Hutchinson, 3 Keb. 785, 84 Eng. Rep. 1011 (1677).

From this report, all that we can tell about the court’s thinking is that it found no convincing reason to grant bail, as was typical in murder cases.3 The rest of the report concerns claims by an attorney. We are told that he contested the jurisdiction of the commission before which Hutchinson was to be tried, apparently a special commission that would have issued pursuant to a statute enacted under Henry VIII.4 The commission lacked jurisdiction, the attorney seemed to suggest, because the crime had occurred in Portugal and thus “out of the Kings Dominions.” The attorney claimed that jurisdiction lay instead with “a Constable and Marshal”–an apparent reference to the High Court of Chivalry, which dealt with treason and murder committed abroad.5 But what, if anything, did the King’s Bench make of the attorney’s jurisdictional claims? And more to the point, what happened after bail was denied? The bail report does not say.

If Hutchinson did ultimately appear before the Court of Chivalry–and if that court accepted a plea of prior acquittal in Portugal–this would be paltry evidence of any common-law principle, which is what Gamble cites Hutchinson to establish. After all, the High Court of Chivalry was a civil-law court prohibited from proceeding under the common law (unlike every other English court of the time save Admiralty). 8 Ric. 2 ch. 5; see also Squibb 162; id., at xxv-xxvi (“The essential distinction between the Court of Chivalry and other courts is . . . that it administers justice in relation to those military matters which are not governed by the common law”). Nor would it be any surprise that we have no report of the proceeding; in fact, “[t]here is no report of a case in which a judge of the Court [of Chivalry] has set out the reasons for his decision earlier than the [20th] century.” Id., at 162.

In the end, we have only two early accounts from judges of what finally became of Hutchinson, and both are indirect and shaky. First, they appear in the reports of cases decided in the Court of Chancery more than a half century after Hutchinson. Second, both judges cite only one source, and it is of lower authority than their own: namely, an account of Hutchinson given by an interested party (a defendant) in a previous, non-criminal case–an account on which the court in that case did not rely or even comment.6 Insofar as our two judges seem to add their own details to the Hutchinson saga, we are not told where they obtained this information or whether it reflects mere guesses as to how gaps in the story should be filled in, decades after the fact. Finally, the two judges’ accounts are not entirely consistent. Still, they are the only early judicial glosses on Hutchinson that we have, so we will work with them.

The more extensive account appears in the case of Gage v. Bulkeley, Ridg. T. H. 263, 27 Eng. Rep. 824 (Ch. 1744), and what the court said there–far from supporting Gamble’s argument–cuts against it. Gage involved a bill in chancery for an account of money deposited with a banker in Paris. The defendants pleaded, as a bar to this lawsuit, “a sentence” “given upon” the same demand in a French court. Ibid. In addressing this plea, Lord Chancellor Hardwicke first determined that foreign judgments are not binding in an English court of law. Here his reasoning was very similar to that found in our dual-sovereignty decisions. Because each judgment rests on the authority of a particular sovereign, the Chancellor thought, it cannot bind foreign courts, which operate by the power of a different sovereign. Id., at 263-264, 27 Eng. Rep., at 824.

Turning next to courts of equity, the Lord Chancellor saw no reason that the rule should be any different; there too, he thought, a foreign judgment is not binding. Id., at 273, 27 Eng. Rep., at 827. But he did allow that in equity a foreign judgment could serve as “evidence, which may affect the right of [a plaintiff] when the cause comes to be heard.” Ibid.

Elaborating on why foreign judgments did not bind English courts, whether of law or equity, the Lord Chancellor explained why Hutchinson was “no proof” to the contrary. In the Chancellor’s telling, Hutchinson was not indicted by the Court of King’s Bench, which could have tried a murder committed in England,7 because that court had no jurisdiction over a homicide committed in Portugal. Gage, Ridge. T. H., at 271, 27 Eng. Rep., at 826-827. Instead, Hutchinson was (as the bail decision indicates) before that court on a writ of habeas corpus, and his case “was referred to the judges to know whether a commission should issue” under a statute similar to the one mentioned in the bail decision. Ibid., 27 Eng. Rep., at 827; see 33 Hen. 8 ch. 28 (1541-1542).8 “And,” he explained, “the judges very rightly and mercifully thought not, because he had undergone one trial already.” Gage, Ridg. T. H., at 271-272, 27 Eng. Rep., at 827 (emphasis added). This suggests that Hutchinson was spared retrial as a matter of discretion (“merc[y]”)–which must be true if the Chancellor was right that foreign judgments were not binding. Indeed, at least one modern scholar agrees (on other grounds as well) that the result in Hutchinson may have been based on “expediency rather than law.” M. Friedland, Double Jeopardy 362-363 (1969).

In the end, then, Gage is doubly damaging to Gamble. First, it squarely rejects the proposition that a litigant in an English court–even a civil litigant in equity–had a right to the benefit of a foreign judgment, a right that the Fifth Amendment might have codified. And second, Gage undermines Gamble’s chief historical example, Hutchinson, by giving a contrary reading of that case–and doing so, no less, in one of the only two judicial accounts of Hutchinson that we have from before the Fifth Amendment.

The other account appears in Burrows v. Jemino, 2 Str. 733, 93 Eng. Rep. 815 (K. B. 1726).9 In Burrows, a party that was sued in England on a bill of exchange sought an injunction against this suit in the Court of Chancery, contending that the suit was barred by the judgment of a court in Italy. In explaining why he would grant the injunction, Lord Chancellor King cited Hutchinson, which he thought had involved an acquittal in Spanish court that was “allowed to be a good bar to any proceedings here.” 2 Str., at 733, 93 Eng. Rep., at 815. This remark, showing that at least one English judge before the founding saw Hutchinson as Gamble does, provides a modicum of support for Gamble’s argument. But that support softens just a few lines down in the report, where the Chancellor discusses the status of foreign judgments in courts of law in particular (as distinct from courts of equity like his own)–i.e., the courts that actually applied the common-law rules later codified by the Fifth Amendment. Here the Chancellor explained that while he personally would have accepted an Italian judgment as barring any suit at law, “other Judges might be of a different opinion.” Ibid. As a whole, then, the Chancellor’s comments in Burrows can hardly be cited to prove that the common law had made up its mind on this matter; just the opposite.

Gamble’s other cases have even less force. The “most instructive” case, he claims, see Brief for Petitioner 13, is the 1775 case of King v. Roche, 1 Leach 134,10 168 Eng. Rep. 169 (K.B.), but that is a curious choice since the Roche court does not so much as mention Hutchinson or even tacitly affirm its supposed holding. The defendant in Roche entered two pleas: prior acquittal abroad and not guilty of the charged crime. All that the Roche court held was that, as a procedural matter, it made no sense to charge the jury with both pleas at once, because a finding for Roche on the first (prior acquittal) would, if successful, bar consideration of the second (not guilty). Roche, 1 Leach, at 135, 168 Eng. Rep., at 169. But on our key question–whether a plea based on a foreign acquittal could be successful–the Roche court said absolutely nothing; it had no occasion to do so. Before the prosecution could reply to Roche’s plea of prior acquittal, he withdrew it, opting for a full trial. The name Hutchinson does not appear even in the marginalia of the 1789 edition of Roche, which existed in 1791. See Captain Roche’s Case, 1 Leach at 138-139.

Hutchinson is mentioned in connection with Roche only after the Fifth Amendment’s ratification, and only in a compiler’s annotation to the 1800 edition of the Roche case report. See 168 Eng. Rep., at 169, n. (a). That annotation in turn cites one case as support for its reading of HutchinsonBeak v. Thyrwhit, 3 Mod. 194, 87 Eng. Rep. 124 (K. B. 1688). But Beak did not involve a foreign prosecution; indeed, it did not involve a prosecution at all. It was an admiralty case for trover and conversion of a ship, and–more to the point–Hutchinson is discussed only in the defendant’s argument in that case, not the court’s response. A report relaying the actual decision in Beak shows that the court ultimately said nothing about the defendant’s Hutchinson argument one way or another. See Beake v. Tyrrell, 1 Show. K. B. 6, 89 Eng. Rep. 411 (1688).11 This same defendant’s argument was the only source of information about Hutchinson on which the Chancellors in Gage and Burrows explicitly relied, as we noted above. All later accounts of Hutchinson seem to stem from this one shallow root.

The last of Gamble’s five pre-Fifth Amendment cases, Rex v. Thomas, 1 Lev. 118, 83 Eng. Rep. 326 (K. B. 1664), did not even involve a foreign prosecution. The defendant was indicted for murder in England, and he pleaded a prior acquittal by a Welsh court. But Wales was then part of the “kingdom of England”; its laws were “the laws of England and no other.” 1 Blackstone 94-95; see Thomas, 1 Lev., at 118, 83 Eng. Rep., at 326-327. So the prior trial in Thomas was not under another sovereign’s laws, making it totally irrelevant for present purposes.

Summing up the import of the preratification cases on which Gamble’s argument rests, we have the following: (1) not a single reported case in which a foreign acquittal or conviction barred a later prosecution for the same act in either Britain or America; (2) not a single reported decision in which a foreign judgment was held to be binding in a civil case in a court of law; (3) fragmentary and not entirely consistent evidence about a 17th-century case in which a defendant named Hutchinson, having been tried and acquitted for murder someplace in the Iberian Peninsula, is said to have been spared a second trial for this crime on some ground, perhaps out of “merc[y],” not as a matter of right; (4) two cases (one criminal, one in admiralty) in which a party invoked a prior foreign judgment, but the court did not endorse or rest anything on the party’s reliance on that judgment; and (5) two Court of Chancery cases actually holding that foreign judgments were not (or not generally) treated as barring trial at common law. This is the flimsy foundation in case law for Gamble’s argument that when the Fifth Amendment was ratified, it was well understood that a foreign criminal judgment would bar retrial for the same act.

Surveying the pre-Fifth Amendment cases in 1959, we concluded that their probative value was “dubious” due to “confused and inadequate reporting.” Bartkus, 359 U. S., at 128, n. 9. Our assessment was accurate then, and the passing years have not made those early cases any clearer or more valuable.


     Not to worry, Gamble responds: Whatever the English courts actually did prior to adoption of the Fifth Amendment, by that time the early English cases were widely thought to support his view. This is a curious argument indeed. It would have us hold that the Fifth Amendment codified a common-law right that existed in legend, not case law. In any event, the evidence that this right was thought to be settled is very thin.

Gamble’s argument is based on treatises, but they are not nearly as helpful as he claims. Alone they do not come close to settling the historical question with enough force to meet Gamble’s particular burden under stare decisis.

Gamble begins with Blackstone, but he reads volumes into a flyspeck. In the body of his Commentaries, all that Blackstone stated was that successive prosecutions could be barred by prior acquittals by “any court having competent jurisdiction of the offence.” 4 Blackstone 335. This is simply a statement of the general double-jeopardy rule, without a word on separate sovereigns. So Gamble directs our attention to a footnote that appears after the phrase “any court having competent jurisdiction.” The footnote refers to the report of Beak v. Thyrwhit, which, as noted, merely rehearses the argument of the defendant in that case, who in turn mentioned Hutchinson–but not in a criminal prosecution, much less one preceded by a foreign trial. This thread tying Blackstone to Hutchinson–a thread woven through footnotes and reports of reports but not a single statement by a court (or even by a party to an actual prosecution)–is tenuous evidence that Blackstone endorsed Gamble’s reading of Hutchinson.

When Gamble’s attorney was asked at argument which other treatises he found most likely to have informed those who ratified the Fifth Amendment, he offered four. See Tr. of Oral Arg. 30-31. But two of the four treatises did not exist when the Fifth Amendment was ratified. See 1 J. Chitty, Criminal Law 458 (1816); 1 T. Starkie, Criminal Pleading 300-301, n. h (1814). And a third discusses not a single case involving a prior prosecution under foreign law. See 2 W. Hawkins, Pleas of the Crown 372 (1739).

That leaves one treatise cited by Gamble that spoke to this issue before ratification, F. Buller, An Introduction to the Law Relative to Trials at Nisi Prius (5th ed. 1788). That treatise concerned the trial of civil cases, id., at 2, and its discussion of prior judgments appeared under the heading “Of Evidence in general,” id., at 221. After considering the evidentiary value of such documents as acts of Parliament, deeds, and depositions, Buller addressed what we would later call issue preclusion. Lifting language from an earlier publication, H. Bathurst, The Theory of Evidence 39 (1761), Buller wrote that a final judgment was “conclusive Evidence” “against all the World” of the factual determinations underlying the judgment. Buller, Nisi Prius, at 245. And it is on this basis that Buller (again lifting from Bathurst) said that even someone acquitted of a crime in Spain “might,” upon indictment in England, “plead the Acquittal in Spain in Bar.” Ibid.

This endorsement of the preclusive effect of a foreign judgment in civil litigation (which even today is not uniformly accepted in this country12) provides no direct support for Gamble since his prior judgment was one of conviction, not acquittal. (There is, after all, a major difference between the preclusive effect of a prior acquittal and that of a prior conviction: Only the first would make a subsequent prosecution pointless, by requiring later courts to assume a defendant’s innocence from the start.) And in any case, the fleeting references in the Buller and Bathurst treatises are hardly sufficient to show that the Members of the First Congress and the state legislators who ratified the Fifth Amendment understood the Double Jeopardy Clause to bar a prosecution in this country after acquittal abroad for the same criminal conduct.

Gamble attempts to augment his support by citing treatises published after the Fifth Amendment was adopted.13 And he notes that the Court in District of Columbia v. Heller, 554 U. S. 570, 605-610 (2008), took treatises of a similar vintage to shed light on the public understanding in 1791 of the right codified by the Second Amendment. But the Heller Court turned to these later treatises only after surveying what it regarded as a wealth of authority for its reading–including the text of the Second Amendment and state constitutions. The 19th-century treatises were treated as mere confirmation of what the Court thought had already been established. Here Gamble’s evidence as to the understanding in 1791 of the double jeopardy right is not at all comparable.


     When we turn from 19th-century treatises to 19th-century state cases, Gamble’s argument appears no stronger. The last time we looked, we found these state cases to be “inconclusive.” Bartkus, 359 U. S., at 131. They seemed to be evenly split and to “manifest conflict[s] in conscience” rather than confident conclusions about the common law. Ibid. Indeed, two of those cases manifested nothing more than a misreading of a then-recent decision of ours. Id., at 130. We see things no differently today.

The distinction between believing successive prosecutions by separate sovereigns unjust and holding them unlawful appears right on the face of the first state case that Gamble discusses. In State v. Brown, 2 N. C. 100, 101 (1794), the court opined that it would be “against natural justice” for a man who stole a horse in the Ohio Territory to be punished for theft in North Carolina just for having brought the horse to that State. To avoid this result, the Brown court simply construed North Carolina’s theft law not to reach the defendant’s conduct. But it did so precisely because the defendant otherwise could face two prosecutions for the same act of theft–despite the common-law rule against double jeopardy for the same “offence”–since “the offence against the laws of this State, and the offence against the laws of [the Ohio Territory] are distinct; and satisfaction made for the offence committed against this State, is no satisfaction for the offence committed against the laws there.” Ibid. Far from undermining the dual-sovereignty rule, Brown expressly affirms it, rejecting outright the idea that a judgment in one sovereign’s court could “be pleadable in bar to an indictment” in another’s. Ibid.

Other state courts were divided. Massachusetts and Michigan courts thought that at least some trials in either federal or state court could bar prosecution in the other, see Commonwealth v. Fuller, 49 Mass. 313, 318 (1844); Harlan v. People, 1 Doug. 207, 212 (Mich. 1843), but those antebellum cases are poor images of the founding-era common law, resting as they do on what we have explained, see Bartkus, 359 U. S., at 130, was a misreading of our then-recent decision in Houston v. Moore, 5 Wheat. 1 (1820), which we discuss below. A Vermont court did take the same view based on its own analysis of the question, State v. Randall, 2 Aik. 89, 100-101 (1827), but just a few years later a Virginia court declared the opposite, Hendrick v. Commonwealth, 32 Va. 707, 713 (1834) (punishment for forgery under both federal and Virginia law is not double punishment for the “same offence” since “the law of Virginia punishes the forgery, not because it is an offence against the U. States, but because it is an offence against this commonwealth”). And South Carolina–a perfect emblem of the time–produced cases cutting both ways. See State v. Antonio, 2 Tread. 776, 781 (1816); State v. Tutt, 2 Bail. 44, 47-48 (1831).

This is not the quantum of support for Gamble’s claim about early American common law that might withstand his burden under stare decisis. And once we look beyond the Nation’s earliest years, the body of state-court decisions appears even less helpful to Gamble’s position. We aptly summarized those cases in Bartkus, 359 U. S., at 134-136, and need not add to that discussion here.14


     Less useful still, for Gamble’s purposes, are the two early Supreme Court cases on which he relies. In the first, a member of the Pennsylvania militia was tried by a state court-martial for the federal offense of deserting the militia. See Houston v. Moore, 5 Wheat. 1 (1820). The accused objected that the state court-martial lacked jurisdiction to try this federal offense. Since the offense could be tried in federal court, the defendant argued, allowing the state court-martial to try him for this crime could expose him to successive federal and state prosecutions for the same offense. Justice Washington answered that a ruling in either federal or state court would bar a second trial in the other. See id., at 31. But as we later explained,

“that language by Mr. Justice Washington reflected his belief that the state statute imposed state sanctions for violation of a federal criminal law. As he viewed the matter, the two trials would not be of similar crimes arising out of the same conduct; they would be of the same crime. Mr. Justice Johnson agreed that if the state courts had become empowered to try the defendant for the federal offense, then such a state trial would bar a federal prosecution. Thus Houston v. Moore can be cited only for the presence of a bar in a case in which the second trial is for a violation of the very statute whose violation by the same conduct has already been tried in the courts of another government empowered to try that question.” Bartkus, 359 U. S., at 130 (citations omitted).

In other words, Justice Washington taught only that the law prohibits two sovereigns (in that case, Pennsylvania and the United States) from both trying an offense against one of them (the United States). That is consistent with our doctrine allowing successive prosecutions for offenses against separate sovereigns. In light of this reading of Houston, the case does not undercut our dual-sovereignty doctrine.

It may seem strange to think of state courts as prosecuting crimes against the United States, but that is just what state courts and commentators writing within a decade of Houston thought it involved. See, e.g.Tutt, 2 Bail., at 47 (“In [Houston], the act punished by the law of the State, was certainly and exclusively an offence against the general Government . . . [whereas h]ere, certainly there is an offence against the State, and a very different one from that committed against the United States” (emphasis added)); 1 J. Kent, Commentaries on American Law 373-374 (1826) (“[M]any . . . acts of [C]ongress . . . permit jurisdiction, over the offences therein described, to be exercised by state magistrates and courts,” and what Houston bars are successive prosecutions for the same “crime against the United States”). Even the scholar Gamble cites for his cause finds Houston not “[o]n point” because it “was discussing the jurisdiction of the state court to try a crime against the nation and impose a fine payable to the latter government.” Grant, Successive Prosecutions by State and Nation: Common Law and British Empire Comparisons, 4 UCLA L. Rev. 1, 7, and n. 27 (1956) (citing Warren, Federal Criminal Laws and the State Courts, 38 Harv. L. Rev. 545 (1925)).

Perhaps feeling Houston wobble, Gamble says pre-emptively that if it is “inconclusive,” Brief for Petitioner 26, other cases are clear. But the other federal case on which he leans is worse for his argument. In United States v. Furlong, 5 Wheat. 184, 197 (1820), we said that an acquittal of piracy in the court of any “civilized State” would bar prosecution in any other nation because piracy, as an “offence within the criminal jurisdiction of all nations,” is “punished by all.”15 Ending his quotation from Furlong at this point, Gamble gives the impression that Furlong rejects any dual-sovereignty rule. But that impression is shattered by the next sentence: “Not so with the crime of murder.” Ibid. As to that crime, the Furlong Court was “inclined to think that an acquittal” in the United States “would not have been a good plea in a Court of Great Britain.” Ibid. (emphasis added). And that was precisely because murder is “punishable under the laws of each State” rather than falling under some “universal jurisdiction.” Ibid. (emphasis added). When it came to crimes that were understood to offend against more than one sovereign, Furlong treated them as separate offenses–just as we have a dozen times since, and just as we do today.

Thus, of the two federal cases that Gamble cites against the dual-sovereignty rule, Houston squares with it and Furlong supports it. Together with the muddle in the early state cases, this undermines Gamble’s claim that the early American bench and bar took the Fifth Amendment to proscribe successive prosecutions by different sovereigns. And without making a splash in the legal practice of the time, a few early treatises by themselves cannot unsettle almost two centuries of precedent.


     Besides appealing to the remote past, Gamble contends that recent changes–one doctrinal, one practical–blunt the force of stare decisis here. They do not.


     If historical claims form the chorus of Gamble’s argument, his refrain is “incorporation.” In Gamble’s telling, the recognition of the Double Jeopardy Clause’s incorporation against the States, see Benton v. Maryland, 395 U. S. 784, 794 (1969), washed away any theoretical foundation for the dual-sovereignty rule, see United States v. Gaudin, 515 U. S. 506, 521 (1995) (abrogating precedent when “subsequent decisions of this Court” have “eroded” its foundations). But this incorporation-changes-everything argument trades on a false analogy.

The analogy Gamble draws is to the evolution of our doctrine on the Fourth Amendment right against unreasonable searches and seizures.16 We have long enforced this right by barring courts from relying on evidence gathered in an illegal search. Thus, in Weeks v. United States, 232 U. S. 383, 391-393 (1914), the Court held that federal prosecutors could not rely on the fruits of an unreasonable search undertaken by federal agents. But what if state or local police conducted a search that would have violated the Fourth Amendment if conducted by federal agents? Before incorporation, the state search would not have violated the Federal Constitution, so federal law would not have barred admission of the resulting evidence in a state prosecution. But by the very same token, under what was termed “the silver-platter doctrine,” state authorities could hand such evidence over to federal prosecutors for use in a federal case. See id., at 398.

Once the Fourth Amendment was held to apply to the States as well as the Federal Government, however, the silver-platter doctrine was scuttled. See Elkins v. United States, 364 U. S. 206 (1960); Wolf v. Colorado, 338 U. S. 25 (1949). Now the fruits of unreasonable state searches are inadmissible in federal and state courts alike.

Gamble contends that the incorporation of the Double Jeopardy Clause should likewise end the dual-sovereignty rule, but his analogy fails. The silver-platter doctrine was based on the fact that the state searches to which it applied did not at that time violate federal law. Once the Fourth Amendment was incorporated against the States, the status of those state searches changed. Now they did violate federal law, so the basis for the silver-platter doctrine was gone. See Elkins, 364 U. S., at 213 (“The foundation upon which the admissibility of state-seized evidence in a federal trial originally rested–that unreasonable state searches did not violate the Federal Constitution–thus disappeared [with incorporation]”).

By contrast, the premises of the dual-sovereignty doctrine have survived incorporation intact. Incorporation meant that the States were now required to abide by this Court’s interpretation of the Double Jeopardy Clause. But that interpretation has long included the dual-sovereignty doctrine, and there is no logical reason why incorporation should change it. After all, the doctrine rests on the fact that only same-sovereign successive prosecutions are prosecutions for the “same offense,” see Part II, supra–and that is just as true after incorporation as before.


     If incorporation is the doctrinal shift that Gamble invokes to justify a departure from precedent, the practical change he cites is the proliferation of federal criminal law. Gamble says that the resulting overlap of federal and criminal codes heightens the risk of successive prosecutions under state and federal law for the same criminal conduct. Thus, Gamble contends, our precedent should yield to ” ‘far-reaching systemic and structural changes’ ” that make our “earlier error all the more egregious and harmful.” South Dakota v. Wayfair, Inc., 585 U. S. ___, ___ (2018) (slip op., at 18). But unlike Gamble’s appeal to incorporation, this argument obviously assumes that the dual-sovereignty doctrine was legal error from the start. So the argument is only as strong as Gamble’s argument about the original understanding of double jeopardy rights, an argument that we have found wanting.

Insofar as the expansion of the reach of federal criminal law has been questioned on constitutional rather than policy grounds, the argument has focused on whether Congress has overstepped its legislative powers under the Constitution. See, e.g.Gonzales v. Raich, 545 U. S. 1, 57-74 (2005) (THOMAS, J., dissenting). Eliminating the dual-sovereignty rule would do little to trim the reach of federal criminal law, and it would not even prevent many successive state and federal prosecutions for the same criminal conduct unless we also overruled the long-settled rule that an “offence” for double jeopardy purposes is defined by statutory elements, not by what might be described in a looser sense as a unit of criminal conduct. See Blockburger v. United States, 284 U. S. 299 (1932). Perhaps believing that two revolutionary assaults in the same case would be too much, Gamble has not asked us to overrule Blockburger along with the dual-sovereignty rule.

*  *  *

     The judgment of the Court of Appeals for the Eleventh Circuit is affirmed.

It is so ordered.

THOMAS, J., concurring

 587 U. S. ____ (2019)

No. 17-646



[June 17, 2019]

JUSTICE THOMAS, concurring.

I agree that the historical record does not bear out my initial skepticism of the dual-sovereignty doctrine. See Puerto Rico v. Sánchez Valle, 579 U. S. ___ (2016) (GINSBURG, J., joined by THOMAS, J. concurring). The founding generation foresaw very limited potential for overlapping criminal prosecutions by the States and the Federal Government.1 The Founders therefore had no reason to address the double jeopardy question that the Court resolves today. Given their understanding of Congress’ limited criminal jurisdiction and the absence of an analogous dual-sovereign system in England, it is difficult to conclude that the People who ratified the Fifth Amendment understood it to prohibit prosecution by a State and the Federal Government for the same offense. And, of course, we are not entitled to interpret the Constitution to align it with our personal sensibilities about ” ‘unjust’ ” prosecutions. Post, at 6 (GINSBURG, J., dissenting); see Currier v. Virginia, 585 U. S. ___, ___ (2018) (plurality opinion) (slip op., at 16) (“While the growing number of criminal offenses in our statute books may be cause for concern, no one should expect (or want) judges to revise the Constitution to address every social problem they happen to perceive” (citation omitted)).

I write separately to address the proper role of the doctrine of stare decisis. In my view, the Court’s typical formulation of the stare decisis standard does not comport with our judicial duty under Article III because it elevates demonstrably erroneous decisions–meaning decisions outside the realm of permissible interpretation–over the text of the Constitution and other duly enacted federal law. It is always “tempting for judges to confuse our own preferences with the requirements of the law,” Obergefell v. Hodges, 576 U. S. ___, ___ (2015) (ROBERTS, C. J., dissenting) (slip op., at 3), and the Court’s stare decisis doctrine exacerbates that temptation by giving the venire of respectability to our continued application of demonstrably incorrect precedents. By applying demonstrably erroneous precedent instead of the relevant law’s text–as the Court is particularly prone to do when expanding federal power or crafting new individual rights–the Court exercises “force” and “will,” two attributes the People did not give it. The Federalist No. 78, p. 465 (C. Rossiter ed. 1961) (capitalization omitted).

We should restore our stare decisis jurisprudence to ensure that we exercise “mer[e] judgment,” ibid., which can be achieved through adherence to the correct, original meaning of the laws we are charged with applying. In my view, anything less invites arbitrariness into judging.2


     The Court currently views stare decisis as a ” ‘principle of policy’ ” that balances several factors to decide whether the scales tip in favor of overruling precedent. Citizens United v. Federal Election Comm’n, 558 U. S. 310, 363 (2010) (quoting Helvering v. Hallock, 309 U. S. 106, 119 (1940)). Among these factors are the “workability” of the standard, “the antiquity of the precedent, the reliance interests at stake, and of course whether the decision was well reasoned.” Montejo v. Louisiana, 556 U. S. 778, 792-793 (2009). The influence of this last factor tends to ebb and flow with the Court’s desire to achieve a particular end, and the Court may cite additional, ad hoc factors to reinforce the result it chooses. But the shared theme is the need for a “special reason over and above the belief that a prior case was wrongly decided” to overrule a precedent. Planned Parenthood of Southeastern Pa. v. Casey, 505 U. S. 833, 864 (1992). The Court has advanced this view of stare decisis on the ground that “it promotes the evenhanded, predictable, and consistent development of legal principles” and “contributes to the actual and perceived integrity of the judicial process.” Payne v. Tennessee, 501 U. S. 808, 827 (1991).

This approach to stare decisis might have made sense in a common-law legal system in which courts systematically developed the law through judicial decisions apart from written law. But our federal system is different. The Constitution tasks the political branches–not the Judiciary–with systematically developing the laws that govern our society. The Court’s role, by contrast, is to exercise the “judicial Power,” faithfully interpreting the Constitution and the laws enacted by those branches. Art. III, §1.


     A proper understanding of stare decisis in our constitutional structure requires a proper understanding of the nature of the “judicial Power” vested in the federal courts. That “Power” is–as Chief Justice Marshall put it–the power “to say what the law is” in the context of a particular “case” or “controversy” before the court. Marbury v. Madison, 1 Cranch 137, 177 (1803); Art. III, §2. Phrased differently, the “judicial Power” “is fundamentally the power to decide cases in accordance with law.” Lawson, The Constitutional Case Against Precedent, 17 Harv. J. L. & Pub. Pol’y 23, 26 (1994) (Lawson). It refers to the duty to exercise “judicial discretion” as distinct from “arbitrary discretion.” The Federalist No. 78, at 468, 471.

That means two things, the first prohibitory and the second obligatory. First, the Judiciary lacks “force” (the power to execute the law) and “will” (the power to legislate). Id., at 465 (capitalization omitted). Those powers are vested in the President and Congress, respectively. “Judicial power is never exercised for the purpose of giving effect to the will of the Judge; always for the purpose of giving effect to the will of the Legislature; or, in other words, to the will of the law.” Osborn v. Bank of United States, 9 Wheat. 738, 866 (1824) (Marshall, C. J.). The Judiciary thus may not “substitute [its] own pleasure to the constitutional intentions of the legislature.” The Federalist No. 78, at 468-469.

Second, “judicial discretion” requires the “liquidat[ion]” or “ascertain[ment]” of the meaning of the law. Id., at 467-468; see id., No. 37. At the time of the founding, “to liquidate” meant “to make clear or plain”; “to render unambiguous; to settle (differences, disputes).” Nelson, Stare Decisis and Demonstrably Erroneous Precedents, 87 Va. L. Rev. 1, 13, and n. 35 (2001) (Nelson) (quoting 8 Oxford English Dictionary 1012 (2d ed. 1991); (internal quotation marks omitted)). Therefore, judicial discretion is not the power to “alter” the law; it is the duty to correctly “expound” it. Letter from J. Madison to N. Trist (Dec. 1831), in 9 The Writings of James Madison 477 (G. Hunt ed. 1910) (Writings of Madison).


     This understanding of the judicial power had long been accepted at the time of the founding. But the federalist structure of the constitutional plan had significant implications for the exercise of that power by the newly created Federal Judiciary. Whereas the common-law courts of England discerned and defined many legal principles in the first instance, the Constitution charged federal courts primarily with applying a limited body of written laws articulating those legal principles. This shift profoundly affects the application of stare decisis today.

Stare decisis has its pedigree in the unwritten common law of England. As Blackstone explained, the common law included “[e]stablished customs” and “[e]stablished rules and maxims” that were discerned and articulated by judges. 1 W. Blackstone, Commentaries on the Laws of England 68-69 (1765) (Blackstone). In the common-law system, stare decisis played an important role because “judicial decisions [were] the principal and most authoritative evidence, that [could] be given, of the existence of such a custom as shall form a part of the common law.” Id., at 69. Accordingly, “precedents and rules must be followed, unless flatly absurd or unjust,” because a judge must issue judgments “according to the known laws and customs of the land” and not “according to his private sentiments” or “own private judgment.” Id., at 69-70. In other words, judges were expected to adhere to precedents because they embodied the very law the judges were bound to apply.

“[C]ommon law doctrines, as articulated by judges, were seen as principles that had been discovered rather than new laws that were being made.” 3-4 G. White, The Marshall Court and Cultural Change, 1815-35, History of the Supreme Court of the United States 129 (1988).3 “It was the application of the dictates of natural justice, and of cultivated reason, to particular cases.” 1 J. Kent, Commentaries on American Law 439 (1826) (Kent); see id., at 439-440 (the common law is ” ‘not the product of the wisdom of some one man, or society of men, in any one age; but of the wisdom, counsel, experience, and observation, of many ages of wise and observing men’ “). The common law therefore rested on “unarticulated social processes to mobilize and coordinate knowledge” gained primarily through “the social experience of the many,” rather than the “specifically articulated reason of the few.” T. Sowell, A Conflict of Visions: Ideological Origins of Political Struggles 49, 42 (1987). In other words, the common law was based in the collective, systematic development of the law through reason. See id., at 49-55.

Importantly, however, the common law did not view precedent as unyielding when it was “most evidently contrary to reason” or “divine law.” Blackstone 69-70. The founding generation recognized that a “judge may mistake the law.” Id., at 71; see also 1 Kent 444 (“Even a series of decisions are not always conclusive evidence of what is law”). And according to Blackstone, judges should disregard precedent that articulates a rule incorrectly when necessary “to vindicate the old [rule] from misrepresentation.” Blackstone 70; see also 1 Kent 443 (“If . . . any solemnly adjudged case can be shown to be founded in error, it is no doubt the right and the duty of the judges who have a similar case before them, to correct the error”). He went further: When a “former decision is manifestly absurd or unjust” or fails to conform to reason, it is not simply “bad law,” but “not law” at all. Blackstone 70 (emphasis). This view–that demonstrably erroneous “blunders” of prior courts should be corrected–was accepted by state courts throughout the 19th century. See, e.g., McDowell v. Oyer, 21 Pa. 417, 423 (1853); Guild v. Eager, 17 Mass. 615, 622 (1822).

This view of precedent implies that even common-law judges did not act as legislators, inserting their own preferences into the law as it developed. Instead, consistent with the nature of the judicial power, common-law judges were tasked with identifying and applying objective principles of law–discerned from natural reason, custom, and other external sources–to particular cases. See Nelson 23-27. Thus, the founding generation understood that an important function of the Judiciary in a common-law system was to ascertain what reason or custom required; that it was possible for courts to err in doing so; and that it was the Judiciary’s responsibility to “examin[e] without fear, and revis[e] without reluctance,” any “hasty and crude decisions” rather than leaving “the character of [the] law impaired, and the beauty and harmony of the system destroyed by the perpetuity of error.” 1 Kent 444.

Federal courts today look to different sources of law when exercising the judicial power than did the common-law courts of England. The Court has long held that “[t]here is no federal general common law.” Erie R. Co. v. Tompkins, 304 U. S. 64, 78 (1938). Instead, the federal courts primarily interpret and apply three bodies of federal positive law–the Constitution; federal statutes, rules, and regulations; and treaties.4 That removes most (if not all) of the force that stare decisis held in the English common-law system, where judicial precedents were among the only documents identifying the governing “customs” or “rules and maxims.” Blackstone 68. We operate in a system of written law in which courts need not–and generally cannot–articulate the law in the first instance. See U. S. Const., Art. I, §1 (vesting “[a]ll legislative Powers” in Congress); Art. 1, §7 (describing the bicameralism and presentment process). The Constitution, federal statutes, and treaties are the law, and the systematic development of the law is accomplished democratically. Our judicial task is modest: We interpret and apply written law to the facts of particular cases.

Underlying this legal system is the key premise that words, including written laws, are capable of objective, ascertainable meaning. As I have previously explained, “[m]y vision of the process of judging is unabashedly based on the proposition that there are right and wrong answers to legal questions.” Thomas, Judging, 45 U. Kan. L. Rev. 1, 5 (1996). Accordingly, judicial decisions may incorrectly interpret the law, and when they do, subsequent courts must confront the question when to depart from them.


     Given that the primary role of federal courts today is to interpret legal texts with ascertainable meanings, precedent plays a different role in our exercise of the “judicial Power” than it did at common law. In my view, if the Court encounters a decision that is demonstrably erroneous–i.e., one that is not a permissible interpretation of the text–the Court should correct the error, regardless of whether other factors support overruling the precedent. Federal courts may (but need not) adhere to an incorrect decision as precedent, but only when traditional tools of legal interpretation show that the earlier decision adopted a textually permissible interpretation of the law. A demonstrably incorrect judicial decision, by contrast, is tantamount to making law, and adhering to it both disregards the supremacy of the Constitution and perpetuates a usurpation of the legislative power.


     When faced with a demonstrably erroneous precedent, my rule is simple: We should not follow it. This view of stare decisis follows directly from the Constitution’s supremacy over other sources of law–including our own precedents. That the Constitution outranks other sources of law is inherent in its nature. See A. Amar, America’s Constitution 5 (2005) (explaining that the Constitution is a constitutive document); Kesavan, The Three Tiers of Federal Law, 100 NW.U. L. Rev. 1479, 1499, n. 99 (2006) (arguing that “[i]t is unnecessary for the Constitution to specify that it is superior to other law because it is higher law made by We the People–and the only such law”). The Constitution’s supremacy is also reflected in its requirement that all judicial officers, executive officers, Congressmen, and state legislators take an oath to “support this Constitution.” Art. VI, cl. 3; see also Art. II, §1, cl. 8 (requiring the President to “solemnly swear (or affirm)” to “preserve, protect and defend the Constitution of the United States”). Notably, the Constitution does not mandate that judicial officers swear to uphold judicial precedents. And the Court has long recognized the supremacy of the Constitution with respect to executive action and “legislative act[s] repugnant to” it. Marbury, 1 Cranch, at 177; Youngstown Sheet & Tube Co. v. Sawyer, 343 U. S. 579, 587-589 (1952); see also The Federalist No. 78, at 467 (“No legislative act, therefore, contrary to the Constitution, can be valid”).

The same goes for judicial precedent. The “judicial Power” must be understood in light of “the Constitution’s status as the supreme legal document” over “lesser sources of law.” Lawson, 29-30. This status necessarily limits “the power of a court to give legal effect to prior judicial decisions” that articulate demonstrably erroneous interpretations of the Constitution because those prior decisions cannot take precedence over the Constitution itself. Ibid. Put differently, because the Constitution is supreme over other sources of law, it requires us to privilege its text over our own precedents when the two are in conflict. I am aware of no legitimate reason why a court may privilege a demonstrably erroneous interpretation of the Constitution over the Constitution itself.5

The same principle applies when interpreting statutes and other sources of law: If a prior decision demonstrably erred in interpreting such a law, federal judges should exercise the judicial power–not perpetuate a usurpation of the legislative power–and correct the error. A contrary rule would permit judges to “substitute their own pleasure” for the law. The Federalist No. 78, at 468; see id., at 466 (” ‘[T]here is no liberty if the power of judging be not separated from the legislative and executive powers’ “).

In sum, my view of stare decisis requires adherence to decisions made by the People–that is, to the original understanding of the relevant legal text–which may not align with decisions made by the Court. Accord, Marshall v. Baltimore & Ohio R. Co., 16 How. 314, 343-344 (1854) (Daniel, J., dissenting) (“Wherever the Constitution commands, discretion terminates” because continued adherence to “palpable error” is a “violation of duty, an usurpation”); Commonwealth v. Posey, 8 Va. 109, 116 (1787) (opinion of Tazewell, J.) (“[A]lthough I venerate precedents, I venerate the written law more”). Thus, no ” ‘special justification’ ” is needed for a federal court to depart from its own, demonstrably erroneous precedent. Halliburton Co. v. Erica P. John Fund, Inc., 573 U. S. 258, 266 (2014); see Nelson 62. Considerations beyond the correct legal meaning, including reliance, workability, and whether a precedent “has become well embedded in national culture,” S. Breyer, Making our Democracy Work: A Judge’s View 152 (2010), are inapposite. In our constitutional structure, our role of upholding the law’s original meaning is reason enough to correct course.6


     Although precedent does not supersede the original meaning of a legal text, it may remain relevant when it is not demonstrably erroneous. As discussed, the “judicial Power” requires the Court to clarify and settle–or, as Madison and Hamilton put it, to “liquidate”–the meaning of written laws. The Federalist No. 78, at 468 (“[I]t is the province of the courts to liquidate and fix [the] meaning and operation [of contradictory laws]”); The Federalist No. 37, at 229 (explaining that the indeterminacy of laws requires courts to “liquidat[e] and ascertai[n]” their meaning “by a series of particular discussions and adjudications”). This need to liquidate arises from the inability of human language to be fully unequivocal in every context. Written laws “have a range of indeterminacy,” and reasonable people may therefore arrive at different conclusions about the original meaning of a legal text after employing all relevant tools of interpretation. See Nelson 11, 14. It is within that range of permissible interpretations that precedent is relevant. If, for example, the meaning of a statute has been “liquidated” in a way that is not demonstrably erroneous (i.e., not an impermissible interpretation of the text), the judicial policy of stare decisis permits courts to constitutionally adhere to that interpretation, even if a later court might have ruled another way as a matter of first impression. Of course, a subsequent court may nonetheless conclude that an incorrect precedent should be abandoned, even if the precedent might fall within the range of permissible interpretations. But nothing in the Constitution requires courts to take that step.

Put another way, there is room for honest disagreement, even as we endeavor to find the correct answer. Compare ___ U. S. ___, 358-371 ( ) (THOMAS, J., concurring in judgment) (concluding that the “historical evidence from the framing” supports the view that the First Amendment permitted anonymous speech), with id., at 371-385 (Scalia, J., dissenting) (concluding that the First Amendment does not protect anonymous speech based on a century of practice in the States). Reasonable jurists can apply traditional tools of construction and arrive at different interpretations of legal texts.

“[L]iquidating” indeterminacies in written laws is far removed from expanding or altering them. See Writings of Madison 477 (explaining that judicial decisions cannot “alter” the Constitution, only “expound” it). The original meaning of legal texts “usually . . . is easy to discern and simple to apply.” A. Scalia, Common Law Courts in a Civil-Law System, in A Matter of Interpretation: Federal Courts and the Law 45 (A. Gutmann ed. 1997). And even in difficult cases, that the original meaning is not obvious at first blush does not excuse the Court from diligently pursuing that meaning. Stopping the interpretive inquiry short–or allowing personal views to color it–permits courts to substitute their own preferences over the text. Although the law may be, on rare occasion, truly ambiguous–meaning susceptible to multiple, equally correct legal meanings–the law never “runs out” in the sense that a Court may adopt an interpretation beyond the bounds of permissible construction.7 In that regard, a legal text is not capable of multiple permissible interpretations merely because discerning its original meaning “requires a taxing inquiry.” Pauley v. BethEnergy Mines, Inc., 501 U. S. 680, 707 (1991) (Scalia, J., dissenting).

This case is a good example. The historical record presents knotty issues about the original meaning of the Fifth Amendment, and JUSTICE GORSUCH does an admirable job arguing against our longstanding interpretation of the Double Jeopardy Clause. Although JUSTICE GORSUCH identifies support for his view in several postratification treatises, see post, at 13-15 (dissenting opinion), I do not find these treatises conclusive without a stronger showing that they reflected the understanding of the Fifth Amendment at the time of ratification. At that time, the common law certainly had not coalesced around this view, see ante, at 10-21, and petitioner has not pointed to contemporaneous judicial opinions or other evidence establishing that his view was widely shared. This lack of evidence, coupled with the unique two-sovereign federalist system created by our Constitution, leaves petitioner to rely on a general argument about “liberty.” Ultimately, I am not persuaded that our precedent is incorrect as an original matter, much less demonstrably erroneous.


     Although this case involves a constitutional provision, I would apply the same stare decisis principles to matters of statutory interpretation. I am not aware of any legal (as opposed to practical) basis for applying a heightened version of stare decisis to statutory-interpretation decisions. Statutes are easier to amend than the Constitution, but our judicial duty is to apply the law to the facts of the case, regardless of how easy it is for the law to change. Cf. Clark v. Martinez, 543 U. S. 371, 402 (2005) (THOMAS, J., dissenting) (explaining that “the realities of the legislative process” will “often preclude readopting the original meaning of a statute that we have upset”). Moreover, to the extent the Court has justified statutory stare decisis based on legislative inaction, this view is based on the “patently false premise that the correctness of statutory construction is to be measured by what the current Congress desires, rather than by what the law as enacted meant.” Johnson v. Transportation Agency, Santa Clara Cty., 480 U. S. 616, 671 (1987) (Scalia, J., dissenting). Finally, even if congressional silence could be meaningfully understood as acquiescence, it still falls short of the bicameralism and presentment required by Article I and therefore is not a “valid way for our elected representatives to express their collective judgment.” Nelson 76.


     For the reasons explained above, the Court’s multifactor approach to stare decisis invites conflict with its constitutional duty. Whatever benefits may be seen to inhere in that approach–e.g., “stability” in the law, preservation of reliance interests, or judicial “humility,” Tr. of Oral Arg. 20, 41-42–they cannot overcome that fundamental flaw.

In any event, these oft-cited benefits are frequently illusory. The Court’s multifactor balancing test for invoking stare decisis has resulted in policy-driven, “arbitrary discretion.” The Federalist No. 78, at 471. The inquiry attempts to quantify the unquantifiable and, by frequently sweeping in subjective factors, provides a ready means of justifying whatever result five Members of the Court seek to achieve. See Holder v. Hall, 512 U. S. 874, 943-944 (1994) (THOMAS, J., concurring in judgment) (describing a ” ‘totality of circumstances’ ” test as “an empty incantation–a mere conjurer’s trick”); Lawrence v. Texas, 539 U. S. 558, 577 (2003) (acknowledging that stare decisis is ” ‘a principle of policy and not a mechanical formula’ “); see also Casey, 505 U. S., at 854-856 (invoking the “kind of reliance that would lend a special hardship to the consequences of overruling and add inequity to the cost of repudiation”). These are not legal questions with right and wrong answers; they are policy choices. See, e.g., A. Goldberg, Equal Justice: The Warren Era of the Supreme Court 96 (1971) (“[T]his concept of stare decisis both justifies the overruling involved in the expansion of human liberties during the Warren years and counsels against the future overruling of the Warren Court libertarian decisions”).

Members of this Court have lamented the supposed “uncertainty” created when the Court overrules its precedent. See Franchise Tax Bd. of Cal. v. Hyattante, at ___-___ (BREYER, J., dissenting) (slip op., at 12-13). But see Lawrencesupra, at 577 (asserting that not overruling precedent would “caus[e] uncertainty”). As I see it, we would eliminate a significant amount of uncertainty and provide the very stability sought if we replaced our malleable balancing test with a clear, principled rule grounded in the meaning of the text.

The true irony of our modern stare decisis doctrine lies in the fact that proponents of stare decisis tend to invoke it most fervently when the precedent at issue is least defensible. See, e.g., Holdersupra, at 944-945 (opinion of THOMAS, J.) (“Stare decisis should not bind the Court to an interpretation of the Voting Rights Act that was based on a flawed method of statutory construction from its inception” and that has created “an irreconcilable conflict” between the Act and the Equal Protection Clause and requires “methodically carving the country into racially designated electoral districts”). It is no secret that stare decisis has had a “ratchet-like effect,” cementing certain grievous departures from the law into the Court’s jurisprudence. Goldberg, supra, at 96. Perhaps the most egregious example of this illegitimate use of stare decisis can be found in our “substantive due process” jurisprudence. McDonald v. Chicago, 561 U. S. 742, 811 (2010) (THOMAS, J., concurring in part and concurring in judgment). The Court does not seriously defend the “legal fiction” of substantive due process as consistent with the original understanding of the Due Process Clause. Ibid. And as I have explained before, “this fiction is a particularly dangerous one” because it “lack[s] a guiding principle to distinguish ‘fundamental’ rights that warrant protection from nonfundamental rights that do not.” Ibid. Unfortunately, the Court has doggedly adhered to these erroneous substantive-due-process precedents again and again, often to disastrous ends. See, e.g., Stenberg v. Carhart, 530 U. S. 914, 982 (2000) (THOMAS, J., dissenting) (“The standard set forth in the Casey plurality has no historical or doctrinal pedigree” and “is the product of its authors’ own philosophical views about abortion” with “no origins in or relationship to the Constitution”). Likewise, the Court refuses to reexamine its jurisprudence about the Privileges or Immunities Clause, thereby relegating a ” ‘clause in the constitution’ ” ” ‘to be without effect.’ ” McDonaldsupra, at 813 (quoting Marbury, 1 Cranch, at 174); see Timbs v. Indiana, 586 U. S. ___, ___ (2019) (THOMAS, J., concurring in judgment) (criticizing the Court’s incorporation doctrine through a clause that addresses procedures). No subjective balancing test can justify such a wholesale disregard of the People’s individual rights protected by the Fourteenth Amendment.

*  *  *

     Our judicial duty to interpret the law requires adherence to the original meaning of the text. For that reason, we should not invoke stare decisis to uphold precedents that are demonstrably erroneous. Because petitioner and the dissenting opinions have not shown that the Court’s dual-sovereignty doctrine is incorrect, much less demonstrably erroneous, I concur in the majority’s opinion.

GINSBURG, J., dissenting

 587 U. S. ____ (2019)

No. 17-646



[June, 17, 2019]


Terance Martez Gamble pleaded guilty in Alabama state court to both possession of a firearm by a person convicted of “a crime of violence” and drug possession, and was sentenced to ten years’ imprisonment, all but one year suspended. Apparently regarding Alabama’s sentence as too lenient, federal prosecutors pursued a parallel charge, possession of a firearm by a convicted felon, in violation of federal law. Gamble again pleaded guilty and received nearly three more years in prison.

Had either the Federal Government or Alabama brought the successive prosecutions, the second would have violated Gamble’s right not to be “twice put in jeopardy . . . for the same offence.” U. S. Const., Amdt. 5, cl. 2. Yet the Federal Government was able to multiply Gamble’s time in prison because of the doctrine that, for double jeopardy purposes, identical criminal laws enacted by “separate sovereigns” are different “offence[s].”

I dissent from the Court’s adherence to that misguided doctrine. Instead of “fritter[ing] away [Gamble’s] libert[y] upon a metaphysical subtlety, two sovereignties,” Grant, The Lanza Rule of Successive Prosecutions, 32 Colum. L. Rev. 1309, 1331 (1932), I would hold that the Double Jeopardy Clause bars “successive prosecutions [for the same offense] by parts of the whole USA.” Puerto Rico v. Sánchez Valle, 579 U. S. ___, ___ (2016) (GINSBURG, J., concurring) (slip op., at 2).



     Gamble urges that the Double Jeopardy Clause incorporates English common law. That law, he maintains, recognized a foreign acquittal or conviction as a bar to retrial in England for the same offense. See Brief for Petitioner 11-15. The Court, in turn, strives mightily to refute Gamble’s account of the common law. See ante, at 8-21. This case, however, does not call for an inquiry into whether and when an 18th-century English court would have credited a foreign court’s judgment in a criminal case. Gamble was convicted in both Alabama and the United States, jurisdictions that are not foreign to each other. English court decisions regarding the respect due to a foreign nation’s judgment are therefore inapposite.


     In United States v. Lanza, 260 U. S. 377 (1922), this Court held that “an act denounced as a crime by both national and state sovereignties is an offense against the peace and dignity of both and may be punished by each.” Id., at 382. Decades later, a sharply divided Court reaffirmed this separate-sovereigns doctrine. Abbate v. United States, 359 U. S. 187 (1959); Bartkus v. Illinois, 359 U. S. 121 (1959). I would not cling to those ill-advised decisions.


     Justification for the separate-sovereigns doctrine centers on the word “offence”: An “offence,” the argument runs, is the violation of a sovereign’s law, the United States and each State are separate sovereigns, ergo successive state and federal prosecutions do not place a defendant in “jeopardy . . . for the same offence.” Ante, at 1, 3-4 (internal quotation marks omitted).

This “compact syllogism” is fatally flawed. See Braun, Praying to False Sovereigns: The Rule Permitting Successive Prosecutions in the Age of Cooperative Federalism, 20 Am. J. Crim. L. 1, 25 (1992). The United States and its constituent States, unlike foreign nations, are “kindred systems,” “parts of ONE WHOLE.” The Federalist No. 82, p. 493 (C. Rossiter ed. 1961) (A. Hamilton). They compose one people, bound by an overriding Federal Constitution. Within that “WHOLE,” the Federal and State Governments should be disabled from accomplishing together “what neither government [could] do alone–prosecute an ordinary citizen twice for the same offence.” Amar & Marcus, Double Jeopardy Law After Rodney King, 95 Colum. L. Rev. 1, 2 (1995).

The notion that the Federal Government and the States are separate sovereigns overlooks a basic tenet of our federal system. The doctrine treats governments as sovereign, with state power to prosecute carried over from years predating the Constitution. See Heath v. Alabama, 474 U. S. 82, 89 (1985) (citing Lanza, 260 U. S., at 382). In the system established by the Federal Constitution, however, “ultimate sovereignty” resides in the governedArizona State Legislature v. Arizona Independent Redistricting Comm’n, 576 U. S. ___, ___ (2015) (slip op., at 31); Martin v. Hunter’s Lessee, 1 Wheat. 304, 324-325 (1816); Braun, supra, at 26-30. Insofar as a crime offends the “peace and dignity” of a sovereign, Lanza, 260 U. S., at 382, that “sovereign” is the people, the “original fountain of all legitimate authority,” The Federalist No. 22, at 152 (A. Hamilton); see Note, Double Prosecution by State and Federal Governments: Another Exercise in Federalism, 80 Harv. L. Rev. 1538, 1542 (1967). States may be separate, but their populations are part of the people composing the United States.

In our “compound republic,” the division of authority between the United States and the States was meant to operate as “a double security [for] the rights of the people.” The Federalist No. 51, at 323 (J. Madison); see Bond v. United States, 564 U. S. 211, 221 (2011). The separate-sovereigns doctrine, however, scarcely shores up people’s rights. Instead, it invokes federalism to withhold liberty. See Bartkus, 359 U. S., at 155-156 (Black, J., dissenting).1

It is the doctrine’s premise that each government has–and must be allowed to vindicate–a distinct interest in enforcing its own criminal laws. That is a peculiar way to look at the Double Jeopardy Clause, which by its terms safeguards the “person” and restrains the government. See, e.g., id., at 155; United States v. All Assets of G.P.S. Automotive Corp., 66 F. 3d 483, 498 (CA2 1995) (Calabresi, J., concurring). The Double Jeopardy Clause embodies a principle, “deeply ingrained” in our system of justice,

“that the State with all its resources and power should not be allowed to make repeated attempts to convict an individual for an alleged offense, thereby subjecting him to embarrassment, expense and ordeal and compelling him to live in a continuing state of anxiety and insecurity, as well as enhancing the possibility that even though innocent he may be found guilty.” Green v. United States, 355 U. S. 184, 187-188 (1957).

“Looked at from the standpoint of the individual who is being prosecuted,” the liberty-denying potential of successive prosecutions, when Federal and State Governments prosecute in tandem, is the same as it is when either prosecutes twice. Bartkus, 359 U. S., at 155 (Black, J., dissenting).


     I turn, next, to further justifications the Court has supplied for the separate-sovereigns doctrine. None should survive close inspection.


     One rationale emphasizes that the Double Jeopardy Clause originally restrained only the Federal Government and did not bar successive state prosecutions. Id., at 124; Lanza, 260 U. S., at 382; Fox v. Ohio, 5 How. 410, 434-435 (1847). Incorporation of the Clause as a restraint on action by the States, effected in Benton v. Maryland, 395 U. S. 784 (1969), has rendered this rationale obsolete.


     Another justification is precedent. In adopting and reaffirming the separate-sovereigns doctrine, the Court relied on dicta from 19th-century opinions. See Abbate, 359 U. S., at 190-193; Bartkus, 359 U. S., at 129-132; Lanza, 260 U. S., at 382-384. The persuasive force of those opinions is diminished by their dubious reasoning. See supra, at 2-4. While drawing upon dicta from prior opinions, the Court gave short shrift to contrary authority. See Braun, supra, at 20-23.

First, the Framers of the Bill of Rights voted down an amendment that would have permitted the Federal Government to reprosecute a defendant initially tried by a State. 1 Annals of Cong. 753 (1789); J. Sigler, Double Jeopardy: The Development of a Legal and Social Policy 30-31 (1969). But cf. ante, at 4-5. Nevermind that this amendment failed; the Court has attributed to the Clause the very meaning the First Congress refrained from adopting.2

Second, early American courts regarded with disfavor the prospect of successive prosecutions by the Federal and State Governments. In Houston v. Moore, 5 Wheat. 1 (1820), Justice Washington expressed concern that such prosecutions would be “very much like oppression, if not worse”; he noted that an acquittal or conviction by one sovereign “might be pleaded in bar of the prosecution before the other.” Id., at 23, 31. The Court today follows Bartkus in distinguishing Justice Washington’s opinion as addressing only the “strange” situation in which a State has prosecuted an offense “against the United States.” Ante, at 24; see Bartkus, 359 U. S., at 130. The distinction is thin, given the encompassing language in Justice Washington’s opinion. Justice Story’s dissent, moreover, declared successive prosecutions for the same offense contrary to “the principles of the common law, and the genius of our free government.” Houston, 5 Wheat., at 72.

Most of the early state decisions cited by the parties regarded successive federal-state prosecutions as unacceptable. See Bartkus, 359 U. S., at 158-159 (Black, J., dissenting). Only one court roundly endorsed a separate-sovereigns theory. Hendrick v. Commonwealth, 32 Va. 707, 713 (1834). The Court reads the state-court opinions as “distin[guishing] between believing successive prosecutions by separate sovereigns unjust and holding them unlawful.” Ante, at 21. I would not read the Double Jeopardy Clause to tolerate “unjust” prosecutions and believe early American courts would have questioned the Court’s distinction. See State v. Brown, 2 N. C. 100, 101 (1794) (allowing successive prosecutions would be “against natural justice, and therefore I cannot believe it to be law”).


     Finally, the Court has reasoned that the separate-sovereigns doctrine is necessary to prevent either the Federal Government or a State from encroaching on the other’s law enforcement prerogatives. Without this doctrine, the Court has observed, the Federal Government, by prosecuting first, could bar a State from pursuing more serious charges for the same offense, Bartkus, 359 U. S., at 137; and conversely, a State, by prosecuting first, could effectively nullify federal law, Abbate, 359 U. S., at 195. This concern envisions federal and state prosecutors working at cross purposes, but cooperation between authorities is the norm. See Bartkus, 359 U. S., at 123. And when federal-state tension exists, successive prosecutions for the federal and state offenses may escape double-jeopardy blockage under the test prescribed in Blockburger v. United States, 284 U. S. 299 (1932). Offenses are distinct, Blockburger held, if “each . . . requires proof of a fact which the other does not.” Id., at 304; see Amar, 95 Colum. L. Rev., at 45-46 (violation of federal civil rights law and state assault law are different offenses).


     The separate-sovereigns doctrine, I acknowledge, has been embraced repeatedly by the Court. But “[s]tare decisis is not an inexorable command.” Payne v. Tennessee, 501 U. S. 808, 828 (1991). Our adherence to precedent is weakest in cases “concerning procedural rules that implicate fundamental constitutional protections.” Alleyne v. United States, 570 U. S. 99, 116, n. 5 (2013). Gamble’s case fits that bill. I would lay the “separate-sovereigns” rationale to rest for the aforesaid reasons and those stated below.


     First, Benton v. Maryland, 395 U. S. 784, which rendered the double jeopardy safeguard applicable to the States, left the separate-sovereigns doctrine the sort of “legal last-man-standing for which we sometimes depart from stare decisis.” Kimble v. Marvel Entertainment, LLC, 576 U. S. ___, ___ (2015) (slip op., at 11). In adopting and cleaving to the doctrine, the Court stressed that originally, the Clause restrained only federal, not state, action. E.g., Bartkus, 359 U. S., at 127; Lanza, 260 U. S., at 382; cf. Abbate, 359 U. S., at 190.

Before incorporation, the separate-sovereigns doctrine had a certain logic: Without a carve-out for successive prosecutions by separate sovereigns, the Double Jeopardy Clause would have barred the Federal Government from prosecuting a defendant previously tried by a State, but would not have prevented a State from prosecuting a defendant previously tried by the Federal Government. Incorporation changed this. Operative against the States since 1969, when the Court decided Benton v. Maryland, 395 U. S. 784, the double jeopardy proscription now applies to the Federal Government and the States alike. The remaining office of the separate-sovereigns doctrine, then, is to enable federal and state prosecutors, proceeding one after the other, to expose defendants to double jeopardy.

The separate-sovereigns doctrine’s persistence contrasts with the fate of analogous dual-sovereignty doctrines following application of the rights at issue to the States. Prior to incorporation of the Fourth Amendment as a restraint on state action, federal prosecutors were free to use evidence obtained illegally by state or local officers, then served up to federal officers on a “silver platter.” See Elkins v. United States, 364 U. S. 206, 208-214 (1960); Weeks v. United States, 232 U. S. 383, 398 (1914). Once the Fourth Amendment applied to the States, abandonment of this “silver platter doctrine” was impelled by “principles of logic” and the reality that, from the perspective of the victim of an unreasonable search and seizure, it mattered not at all “whether his constitutional right ha[d] been invaded by a federal agent or by a state officer.” Elkins, 364 U. S., at 208, 215. As observed by Justice Harlan, Elkins‘ abandonment of a separate-sovereigns exception to the exclusionary rule was at odds with retention of the separate-sovereigns doctrine for double jeopardy purposes in Abbate and Bartkus. See 364 U. S., at 252.

Similarly, before incorporation of the Fifth Amendment privilege against self-incrimination, the Court held that the privilege did not prevent state authorities from compelling a defendant to provide testimony that could incriminate him or her in another jurisdiction. Knapp v. Schweitzer, 357 U. S. 371, 375-381 (1958). After application of the self-incrimination privilege to the States, the Court concluded that its prior position was incompatible with the “policies and purposes” of the privilege. Murphy v. Waterfront Comm’n of N. Y. Harbor, 378 U. S. 52, 55, 77 (1964). No longer, the Court held, could a witness ” 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 (internal quotation marks omitted; emphasis added).

The Court regards incorporation as immaterial because application of the Double Jeopardy Clause to the States did not affect comprehension of the word “offence” to mean the violation of one sovereign’s law. Ante, at 28. But the Court attributed a separate-sovereigns meaning to “offence” at least in part because the Double Jeopardy Clause did not apply to the States. See supra, at 5. Incorporation of the Clause should prompt the Court to consider the protection against double jeopardy from the defendant’s perspective and to ask why each of two governments within the United States should be permitted to try a defendant once for the same offense when neither could try him or her twice.


     The expansion of federal criminal law has exacerbated the problems created by the separate-sovereigns doctrine. Ill effects of the doctrine might once have been tempered by the limited overlap between federal and state criminal law. All Assets of G.P.S. Automotive, 66 F. 3d, at 498 (Calabresi, J., concurring). In the last half century, however, federal criminal law has been extended pervasively into areas once left to the States. Guerra, The Myth of Dual Sovereignty: Multijurisdictional Drug Law Enforcement and Double Jeopardy, 73 N. C. L. Rev. 1159, 1165-1192 (1995); Brief for Sen. Orrin Hatch as Amicus Curiae 8-14. This new “age of ‘cooperative federalism,’ [in which] the Federal and State Governments are waging a united front against many types of criminal activity,” Murphy, 378 U. S., at 55-56, provides new opportunities for federal and state prosecutors to “join together to take a second bite at the apple,” All Assets of G.P.S. Automotive, 66 F. 3d, at 498 (Calabresi, J., concurring).3 This situation might be less troublesome if successive prosecutions occurred only in “instances of peculiar enormity, or where the public safety demanded extraordinary rigor.” Fox, 5 How., at 435. The run-of-the-mill felon-in-possession charges Gamble encountered indicate that, in practice, successive prosecutions are not limited to exceptional circumstances.


     Against all this, there is little to be said for keeping the separate-sovereigns doctrine. Gamble’s case “do[es] not implicate the reliance interests of private parties.” Alleyne, 570 U. S., at 119 (SOTOMAYOR, J., concurring). The closest thing to a reliance interest would be the interest Federal and State Governments have in avoiding avulsive changes that could complicate ongoing prosecutions. As the Court correctly explains, however, overruling the separate-sovereigns doctrine would not affect large numbers of cases. See ante, at 28-29. In prosecutions based on the same conduct, federal and state prosecutors will often charge offenses having different elements, charges that, under Blockburger, will not trigger double jeopardy protection. See Poulin, Double Jeopardy Protection From Successive Prosecution: A Proposed Approach, 92 Geo. L. J. 1183, 1244-1245 (2004); Brief for Criminal Defense Experts as Amici Curiae 5-11.4

Notably, the Federal Government has endeavored to reduce the incidence of “same offense” prosecutions. Under the Petite policy adopted by the Department of Justice,5 the Department will pursue a federal prosecution “based on substantially the same act(s) or transaction(s)” previously prosecuted in state court only if the first prosecution left a “substantial federal interest . . . demonstrably unvindicated” and a Department senior official authorizes the prosecution. Dept. of Justice, Justice Manual §9-2.031(A) (rev. July 2009).

At oral argument, the Government estimated that it authorizes only “about a hundred” Petite prosecutions per year. Tr. of Oral Arg. 54. But see id., at 65-66 (referring to the “few hundred successive prosecutions that [the Government] bring[s] each year”). Some of these prosecutions will not implicate double jeopardy, as the Petite policy uses a same-conduct test that is broader than the Blockburger same-elements test. And more than half the States forbid successive prosecutions for all or some offenses previously resolved on the merits by a federal or state court. Brief for Criminal Defense Experts as Amici Curiae 4-5, and n. 2 (collecting statutes); Brief for State of Texas et al. as Amici Curiae 28-30, and nn. 6-15 (same). In short, it is safe to predict that eliminating the separate-sovereigns doctrine would spark no large disruption in practice.

*  *  *

     The separate-sovereigns doctrine, especially since Bartkus and Abbate, has been subject to relentless criticism by members of the bench, bar, and academy. Nevertheless, the Court reaffirms the doctrine, thereby diminishing the individual rights shielded by the Double Jeopardy Clause. Different parts of the “WHOLE” United States should not be positioned to prosecute a defendant a second time for the same offense. I would reverse Gamble’s federal conviction.

GORSUCH, J., dissenting

 587 U. S. ____ (2019)

No. 17-646



[June 17, 2019]

JUSTICE GORSUCH, dissenting.

A free society does not allow its government to try the same individual for the same crime until it’s happy with the result. Unfortunately, the Court today endorses a colossal exception to this ancient rule against double jeopardy. My colleagues say that the federal government and each State are “separate sovereigns” entitled to try the same person for the same crime. So if all the might of one “sovereign” cannot succeed against the presumptively free individual, another may insist on the chance to try again. And if both manage to succeed, so much the better; they can add one punishment on top of the other. But this “separate sovereigns exception” to the bar against double jeopardy finds no meaningful support in the text of the Constitution, its original public meaning, structure, or history. Instead, the Constitution promises all Americans that they will never suffer double jeopardy. I would enforce that guarantee.


     “Fear and abhorrence of governmental power to try people twice for the same conduct is one of the oldest ideas found in western civilization.”1 Throughout history, people have worried about the vast disparity of power between governments and individuals, the capacity of the state to bring charges repeatedly until it wins the result it wants, and what little would be left of human liberty if that power remained unchecked. To address the problem, the law in ancient Athens held that “[a] man could not be tried twice for the same offense.”2 The Roman Republic and Empire incorporated a form of double jeopardy protection in their laws.3 The Old Testament and later church teachings endorsed the bar against double jeopardy too.4 And from the earliest days of the common law, courts recognized that to “punish a man twice over for one offence” would be deeply unjust.5

The rule against double jeopardy was firmly entrenched in both the American colonies and England at the time of our Revolution.6 And the Fifth Amendment, which prohibits placing a defendant “twice . . . in jeopardy of life or limb” for “the same offence” sought to carry the traditional common law rule into our Constitution.7 As Joseph Story put it, the Constitution’s prohibition against double jeopardy grew from a “great privilege secured by the common law” and meant “that a party shall not be tried a second time for the same offence, after he has once been convicted, or acquitted of the offence charged, by the verdict of a jury, and judgment has passed thereon for or against him.”8

Given all this, it might seem that Mr. Gamble should win this case handily. Alabama prosecuted him for violating a state law that “prohibits a convicted felon from possessing a pistol” and sentenced him to a year in prison.9 But then the federal government, apparently displeased with the sentence, charged Mr. Gamble under 18 U. S. C. §922(g)(1) with being a felon in possession of a firearm based on the same facts that gave rise to the state prosecution. Ultimately, a federal court sentenced him to 46 months in prison and three years of supervised release. Most any ordinary speaker of English would say that Mr. Gamble was tried twice for “the same offence,” precisely what the Fifth Amendment prohibits. Tellingly, no one before us doubts that if either the federal government or Alabama had prosecuted Mr. Gamble twice on these facts and in this manner, it surely would have violated the Constitution.

So how does the government manage to evade the Fifth Amendment’s seemingly plain command? On the government’s account, the fact that federal and state authorities split up the prosecutions makes all the difference. Though the Double Jeopardy Clause doesn’t say anything about allowing “separate sovereigns” to do sequentially what neither may do separately, the government assures us the Fifth Amendment’s phrase “same offence” does this work. Adopting the government’s argument, the Court supplies the following syllogism: “[A]n ‘offence’ is defined by a law, and each law is defined by a sovereign. So where there are two sovereigns, there are two laws, and two ‘offences.’ ” Ante, at 3-4.

But the major premise of this argument–that “where there are two laws there are ‘two offenses’ “–is mistaken. We know that the Constitution is not so easily evaded and that two statutes can punish the same offense.10 The framers understood the term “offence” to mean a “transgression.”11 And they understood that the same transgression might be punished by two pieces of positive law: After all, constitutional protections were not meant to be flimsy things but to embody “principles that are permanent, uniform, and universal.”12 As this Court explained long ago in Blockburger v. United States, “where the same act or transaction constitutes a violation of two distinct statutory provisions, the test to be applied to determine whether there are two offenses or only one, is whether each provision requires proof of a fact which the other does not.”13 So if two laws demand proof of the same facts to secure a conviction, they constitute a single offense under our Constitution and a second trial is forbidden. And by everyone’s admission, that is exactly what we have here: The statute under which the federal government proceeded required it to prove no facts beyond those Alabama needed to prove under state law to win its conviction; the two prosecutions were for the same offense.

That leaves the government and the Court to rest on the fact that distinct governmental entities, federal and state, enacted these identical laws. This, we are told, is enough to transform what everyone agrees would otherwise be the same offense into two different offenses. But where is that distinction to be found in the Constitution’s text or original public understanding? We know that the framers didn’t conceive of the term “same offence” in some technical way as referring only to the same statute. And if double jeopardy prevents one government from prosecuting a defendant multiple times for the same offense under the banner of separate statutory labels, on what account can it make a difference when many governments collectively seek to do the same thing?

The government identifies no evidence suggesting that the framers understood the term “same offence” to bear such a lawyerly sovereign-specific meaning. Meanwhile, Blackstone’s Commentaries explained how “Roman law,” “Athens,” “the Jewish republic,” and “English Law” addressed the singular “offence of homicide,” and how the Roman, Gothic, and ancient Saxon law approached the singular “offence of arson.”14 Other treatises of the period contain similar taxonomies of “offences” that are not sovereign-specific.15 Members of the Continental Congress, too, used the word “offence” in this same way. In 1786, a congressional committee endorsed federal control over import duties because otherwise “thirteen separate authorities” might “ordain various penalties for the same offence.”16 In 1778, the Continental Congress passed a resolution declaring that a person should not be tried in state court “for the same offense, for which he had previous thereto been tried by a Court Martial.”17 And in 1785, the Continental Congress considered an ordinance declaring that a defendant could “plead a formal Acquital on a Trial” in a maritime court “for the same supposed Offences, in a similar Court in one of the other United States.”18 In all of these examples, early legislators–including many of the same people who would vote to add the Fifth Amendment to the Bill of Rights just a few years later–recognized that transgressions of state and federal law could constitute the “same offence.”

The history of the Double Jeopardy Clause itself supplies more evidence yet. The original draft prohibited “more than one trial or one punishment for the same offence.”19 One representative then proposed adding the words “by any law of the United States” after “same offence.”20 That proposal clearly would have codified the government’s sovereign-specific view of the Clause’s operation. Yet, Congress proceeded to reject it.

Viewed from the perspective of an ordinary reader of the Fifth Amendment, whether at the time of its adoption or in our own time, none of this can come as a surprise. Imagine trying to explain the Court’s separate sovereigns rule to a criminal defendant, then or now. Yes, you were sentenced to state prison for being a felon in possession of a firearm. And don’t worry–the State can’t prosecute you again. But a federal prosecutor can send you to prison again for exactly the same thing. What’s more, that federal prosecutor may work hand-in-hand with the same state prosecutor who already went after you. They can share evidence and discuss what worked and what didn’t the first time around. And the federal prosecutor can pursue you even if you were acquitted in the state case. None of that offends the Constitution’s plain words protecting a person from being placed “twice . . . in jeopardy of life or limb” for “the same offence.” Really?


     Without meaningful support in the text of the Double Jeopardy Clause, the government insists that the separate sovereigns exception is at least compelled by the structure of our Constitution. On its view, adopted by the Court today, allowing the federal and state governments to punish the same defendant for the same conduct “honors the substantive differences between the interests that two sovereigns can have” in our federal system. Ante, at 5.

But this argument errs from the outset. The Court seems to assume that sovereignty in this country belongs to the state and federal governments, much as it once belonged to the King of England. But as Chief Justice Marshall explained, “[t]he government of the Union . . . is emphatically, and truly, a government of the people,” and all sovereignty “emanates from them.”21 Alexander Hamilton put the point this way: “[T]he national and State systems are to be regarded” not as different sovereigns foreign to one another but “as ONE WHOLE.”22 Under our Constitution, the federal and state governments are but two expressions of a single and sovereign people.

This principle resonates throughout our history and law. State courts that refused to entertain federal causes of action found little sympathy when attempting the very separate sovereigns theory underlying today’s decision.23 In time, too, it became clear that federal courts may decide state-law issues, and state courts may decide federal questions.24 Even in the criminal context, this Court has upheld removal of some state criminal actions to federal court.25 And any remaining doubt about whether the States and the federal government are truly separate sovereigns was ultimately “resolved by war.”26

From its mistaken premise, the Court continues to the flawed conclusion that the federal and state governments can successively prosecute the same person for the same offense. This turns the point of our federal experiment on its head. When the “ONE WHOLE” people of the United States assigned different aspects of their sovereign power to the federal and state governments, they sought not to multiply governmental power but to limit it. As this Court has explained, “[b]y denying any one government complete jurisdiction over all the concerns of public life, federalism protects the liberty of the individual from arbitrary power.”27 Yet today’s Court invokes federalism not to protect individual liberty but to threaten it, allowing two governments to achieve together an objective denied to each. The Court brushes this concern aside because “the powers of the Federal Government and the States often overlap,” which “often results in two layers of regulation.” Ante, at 10. But the Court’s examples–taxation, alcohol, and marijuana–involve areas that the federal and state governments each may regulate separately under the Constitution as interpreted by this Court. That is miles away from the separate sovereigns exception, which allows the federal and state governments to accomplish together what neither may do separately consistent with the Constitution’s commands. As Justice Black understood, the Court’s view today “misuse[s] and desecrat[es] . . . the concept” of federalism.28 For “it is just as much an affront to . . . human freedom for a man to be punished twice for the same offense” by two parts of the people’s government “as it would be for one . . . to throw him in prison twice for the offense.”29



     If the Constitution’s text and structure do not supply persuasive support for the government’s position, what about a more thorough exploration of the common law from which the Fifth Amendment was drawn?

By 1791 when the Fifth Amendment was adopted, an array of common law authorities suggested that a prosecution in any court, so long as the court had jurisdiction over the offense, was enough to bar future reprosecution in another court. Blackstone, for example, reported that an acquittal “before any court having competent jurisdiction of the offence” could be pleaded “in bar of any subsequent accusation for the same crime.”30 For support, Blackstone pointed to Beak v. Tyrhwhit,31 a 1688 case in which the reporter described an acquittal in a foreign country followed by an attempted second prosecution in England that the court held impermissible. Another treatise by William Hawkins likewise considered it “settled” as early as 1716 “[t]hat an Acquittal in any Court whatsoever, which has a Jurisdiction of the Cause, is as good a Bar of any subsequent Prosecution for the same Crime.”32

What these authorities suggest many more confirm. Henry Bathurst’s 1761 treatise on evidence taught that “a final Determination in a Court having competent Jurisdiction is conclusive in all Courts of concurrent Jurisdiction.”33 Nor was this merely a rule about the competency of evidence, as the next sentence reveals: “If A. having killed a Person in Spain was there prosecuted, tried, and acquitted, and afterwards was indicted here [in England], he might plead the Acquittal in Spain in Bar.”34 Francis Buller’s 1772 treatise repeated the same rule, articulating it the same way.35 And to illustrate their point, both treatises cited the 1678 English case of King v. Hutchinson. Although no surviving written report of Hutchinson remains, several early common law cases–including Beak v. Thyrwhit,36 Burrows v. Jemino,37 and King v. Roche38–described its holding in exactly the same way the treatise writers did: All agreed that it barred the retrial in England of a defendant previously tried for murder in Spain or Portugal.

When they envisioned the relationship between the national government and the States under the new Constitution, the framers sometimes referenced by way of comparison the relationship between Wales, Scotland, and England.39 And prosecutions in one of these places pretty plainly barred subsequent prosecutions for the same offense in the others. So, for example, treatises explained that “an Acquittal of Murder at a Grand Sessions in Wales, may be pleaded to an Indictment for the same Murder in England. For the Rule is, That a Man’s Life shall not be brought into Danger for the same Offence more than once.”40 Indeed, when an English county indicted a defendant “for a murder committed . . . in Wales,” it was barred from proceeding when the court learned that the defendant had already been tried and acquitted “of the same offence” in Wales.41

Against this uniform body of common law weighs Gage v. Bulkeley–a civil, not criminal, case from 1744 that suggested Hutchinson had held only that the English courts lacked jurisdiction to try a defendant for an offense committed in Portugal. Because “the murder was committed in Portugal,” Gage argued, “the Court of King’s Bench could not indict him, and there was no method of trying him but upon a special commission.”42 But no one else–not the treatise writers or the other English cases that favorably cited Hutchinson–adopted Gage‘s restrictive reading of that precedent.

In the end, then, it’s hard to see how anyone consulting the common law in 1791 could have avoided this conclusion: While the issue may not have arisen often, the great weight of authority indicated that successive prosecutions by different sovereigns–even sovereigns as foreign to each other as England and Portugal–were out of bounds. And anyone familiar with the American federal system likely would have thought the rule applied with even greater force to successive prosecutions by the United States and a constituent State, given that both governments derive their sovereignty from the American people.

Unable to summon any useful preratification common law sources of its own, the government is left to nitpick those that undermine its position. For example, the Court dismisses Beak because “Hutchinson is discussed only in the defendant’s argument in that case, not the court’s response.” Ante, at 16. But the Beak court did not reject the Hutchinson argument, and counsel’s use of the case sheds light on how 17th- and 18th-century lawyers understood the double jeopardy bar. The Court likewise derides King v. Thomas as “totally irrelevant” because in the 17th century, Wales and England shared the same laws. But our federal and state governments share the same fundamental law and source of authority, and the Wales example is at least somewhat analogous to our federal system.43 Finally, the Court complains that Roche‘s footnote citing Hutchinson was added only in 1800, after the Fifth Amendment’s ratification. Ante, at 16. But that is hardly a point for the government, because even so it provides an example of a later reporter attempting to describe the pre-existing state of the law; nor, as it turns out, was the footnote even essential to the Roche court’s original analysis and conclusion reached in 1775, well before the Fifth Amendment’s ratification.44 And among all these complaints, we should not lose the forest for the trees. The Court’s attempts to explain away so many uncomfortable authorities are lengthy, detailed, even herculean. But in the end, neither it nor the government has mustered a single preratification common law authority approving a case of successive prosecutions by separate sovereigns for the same offense.


     What we know about the common law before the Fifth Amendment’s ratification in 1791 finds further confirmation in how later legal thinkers in both England and America described the rule they had inherited.

Start with England. As it turns out, “it would have been difficult to have made more than the most cursory examination of nineteenth century or later English treatises or digests without encountering” the Hutchinson rule.45 In 1802, a British treatise explained that “an acquittal on a criminal charge in a foreign country may be pleaded in bar of an indictment for the same offence in England.”46 Three decades later, another treatise observed (citing Hutchinson) that “[a]n acquittal by a competent jurisdiction abroad is a bar to an indictment for the same offence before any other tribunal.”47 In 1846, the Scottish High Court of Justiciary declared that “[i]f a man has been tried for theft in England, we would not try him again here.”48 Twentieth century treatises recited the same rule.49 In 1931, the American Law Institute stated that “[i]f a person has been acquitted in a court of competent jurisdiction for an offense in another country he may not be tried for the same offense again in an English Court.”50 And in 1971, an English judge explained that the bar on “double jeopardy . . . has always applied whether the previous conviction or acquittal based on the same facts was by an English court or by a foreign court.”51 The Court today asks us to assume that all these legal authorities misunderstood the common law’s ancient rule. I would not.

Even more pertinently, consider how 18th-century Americans understood the double jeopardy provision they had adopted. The legal treatises an American lawyer practicing between the founding and the Civil War might have consulted uniformly recited the Hutchinson rule as black letter law. Chancellor Kent wrote that “the plea of autrefois acquit, resting on a prosecution [in] any civilized state, would be a good plea in any other civilized state.”52 Thomas Sergeant explained that “[w]here the jurisdiction of the United States court and of a state Court is concurrent, the sentence of either court, whether of conviction or acquittal, may be pleaded in bar to a prosecution in the other.”53 William Rawle echoed that conclusion in virtually identical words.54 Indeed, one early commentator wrote that a “principal reason” for the Double Jeopardy Clause was to prevent successive state and federal prosecutions, which he considered to be against “[n]atural justice.”55 Nor did these treatises purport to invent a new rule; they claimed only to recite the traditional one.

This Court’s early decisions reflected the same principle. In Houston v. Moore, a Pennsylvania court-martial tried a member of the state militia for desertion under an “act of the legislature of Pennsylvania.”56 The defendant objected that the state court-martial lacked jurisdiction because federal law criminalized the same conduct and prosecuting him in the state court could thus expose him to double jeopardy. In an opinion by Justice Washington, the Court disagreed and allowed the prosecution, but reassured the defendant that “if the jurisdiction of the two Courts be concurrent, the sentence of either Court, either of conviction or acquittal, might be [later] pleaded in bar of the prosecution before the other.”57 In dissent, Justice Story thought the state court lacked jurisdiction because otherwise the defendant would be “liable to be twice tried and punished for the same offence, against the manifest intent of the act of Congress, the principles of the common law, and the genius of our free government.”58 But notice the point of agreement between majority and dissent: Both acknowledged that a second prosecution for the same underlying offense would be prohibited even if brought by a separate government.59

Another case decided the same year also reflected the Hutchinson rule. In United States v. Furlong, one British subject killed another on the high seas, and the killer was indicted in an American federal court for robbery and murder. This Court unanimously held that “[r]obbery on the seas is considered as an offence within the criminal jurisdiction of all nations” that can therefore be “punished by all,” and there can be “no doubt that the plea of autre fois acquit [double jeopardy] would be good in any civilized State, though resting on a prosecution instituted in the Courts of any other civilized State.”60

A number of early state cases followed the same rule. Indeed, the Court today acknowledges that Massachusetts, Michigan, and Vermont all followed HutchinsonAnte, at 22.61 The Court agrees that South Carolina did too,62 but it believes that a later South Carolina case might have deviated from the Hutchinson rule. That decision, however, contains at best only “an inconclusive discussion coming from a State whose highest court had previously stated unequivocally that a bar against double prosecutions would exist.”63

In the face of so much contrary authority, the Court winds up leaning heavily on a single 1794 North Carolina Superior Court decision, State v. Brown. But the Court’s choice here is revealing. True, Brown said that a verdict in North Carolina would not be “pleadable in bar to an indictment preferred against [the defendant] in the Territory South of the Ohio.”64 But the Court leaves out what happened next. Brown went on to reject concurrent jurisdiction because trying the defendant “according to the several laws of each State” could result in him being “cropped in one, branded and whipped in another, imprisoned in a third, and hanged in a fourth; and all for one and the same offence.”65 The North Carolina court viewed that result as “against natural justice” and “therefore [could] not believe it to be law.”66 So it is that the principal support the Court cites for its position is a state case that both (1) regarded transgressions of the laws of a State and a U. S. territory as the “same offence,” and (2) expressed aversion at the thought of both jurisdictions punishing the defendant for that singular offense.67


     With the text, principles of federalism, and history now arrayed against it, the government is left to suggest that we should retain the separate sovereigns exception under the doctrine of stare decisis. But if that’s the real basis for today’s result, let’s at least acknowledge this: By all appearances, the Constitution as originally adopted and understood did not allow successive state and federal prosecutions for the same offense, yet the government wants this Court to tolerate the practice anyway.

Stare decisis has many virtues, but when it comes to enforcing the Constitution this Court must take (and always has taken) special care in the doctrine’s application. After all, judges swear to protect and defend the Constitution, not to protect what it prohibits. And while we rightly pay heed to the considered views of those who have come before us, especially in close cases, stare decisis isn’t supposed to be “the art of being methodically ignorant of what everyone knows.”68 Indeed, blind obedience to stare decisis would leave this Court still abiding grotesque errors like Dred Scott v. Sandford,69 Plessy v. Ferguson,70 and Korematsu v. United States.71 As Justice Brandeis explained, “in cases involving the Federal Constitution, where correction through legislative action is practically impossible, this Court has often overruled its earlier decisions. The Court bows to the lessons of experience and the force of better reasoning, recognizing that the process of trial and error, so fruitful in the physical sciences, is appropriate also in the judicial function.”72

For all these reasons, while stare decisis warrants respect, it has never been ” ‘an inexorable command,’ “73 and it is “at its weakest when we interpret the Constitution.”74 In deciding whether one of our cases should be retained or overruled, this Court has traditionally considered “the quality of the decision’s reasoning; its consistency with related decisions; legal developments since the decision; and reliance on the decision.”75 Each of these factors, I believe, suggests we should reject the separate sovereigns exception.

Take the “quality of [the] reasoning.”76 The first cases to suggest that successive prosecutions by state and federal authorities might be permissible did not seek to address the original meaning of the word “offence,” the troubling federalism implications of the exception, or the relevant historical sources. Between 1847 and 1850, the Court decided a pair of cases, United States v. Marigold77 and Fox v. Ohio.78 While addressing other matters in those decisions, the Court offered passing approval to the possibility of successive state and federal prosecutions, but did so without analysis and without actually upholding a successive conviction. Indeed, in place of a careful constitutional analysis, the Fox Court merely offered its judgment that “the benignant spirit” of prosecutors could be relied on to protect individuals from too many repetitive prosecutions.79 We do not normally give precedential effect to such stray commentary.

Perhaps the first real roots of the separate sovereigns exception can be traced to this Court’s 1852 decision in Moore v. Illinois.80 As it did five years later and more notoriously in Dred Scott,81 the Court in Moore did violence to the Constitution in the name of protecting slavery and slaveowners. In Dred Scott the Court held that the Due Process Clause prevented Congress from prohibiting slavery in the territories, though of course the Clause did nothing of the sort.82 And in Moore the Court upheld a state fugitive slave law that it judged important because the States supposedly needed “to protect themselves against the influx either of liberated or fugitive slaves, and to repel from their soil a population likely to become burdensome and injurious, either as paupers or criminals.”83 The defendant, who had harbored a fugitive slave, objected that upholding the state law could potentially expose him to double prosecutions by the state and federal governments. The Court rejected that argument, reasoning simply that such double punishment could be consistent with the Constitution if the defendant had violated both state and federal law.84 Yet notably, even here, the Court did not actually approve a successive prosecution.

Nor did the trajectory of the separate sovereigns exception improve much from there. The first time the Court actually approved an “instance of double prosecution [and] failed to find some remedy . . . to avoid it” didn’t arrive until 1922.85 In that case, United States v. Lanza,86 the federal government prosecuted the defendants for manufacturing, transporting, and possessing alcohol in violation of the National Prohibition Act. The defendants argued that they had already been prosecuted by the State of Washington for the same offense. But, notably, the defendants did not directly question the permissibility of successive prosecutions for the same offense under state and federal law. Instead, the defendants argued that both of the laws under which they were punished really derived from the “same sovereign:” the national government, by way of the Eighteenth Amendment that authorized Prohibition. After rejecting that argument as an “erroneous view of the matter,” the Court proceeded on, perhaps unnecessarily, to offer its view that “an act denounced as a crime by both national and state sovereignties is an offense against the peace and dignity of both and may be punished by each.”87 Given that the Court was not asked directly to consider the propriety of successive prosecutions under separate state and federal laws for the same offense, it is perhaps unsurprising the Court did not consult the original meaning of the Double Jeopardy Clause or consult virtually any of the relevant historical sources before offering its dictum.

It matters, too, that these cases “were decided by the narrowest of margins, over spirited dissents challenging the basic underpinnings of those decisions.”88 In Moore, Justice McLean wrote that although “the Federal and State Governments emanate from different sovereignties,” they “operate upon the same people, and should have the same end in view.”89 He “deeply regret[ted] that our government should be an exception to a great principle of action, sanctioned by humanity and justice.”90 Bartkus and Abbate, cases decided in the 1950s that more clearly approved the separate sovereigns exception, were decided only by 5-to-4 and 6-to-3 margins, and Justice Black’s eloquent dissents in those cases have triggered an avalanche of persuasive academic support.91

What is more, the “underpinnings” of the separate sovereigns exception have been “erode[d] by subsequent decisions of this Court.”92 When this Court decided MooreLanzaBartkus, and Abbate, the Double Jeopardy Clause applied only to the federal government under this Court’s decision in Palko v. Connecticut.93 In those days, one might have thought, the separate sovereigns exception at least served to level the playing field between the federal government and the States: If a State could retry a defendant after a federal trial, then the federal government ought to be able to retry a defendant after a state trial. But in time the Court overruled Palko and held that the Double Jeopardy Clause does apply to the States–and, with that, a premise once thought important to the exception fell away.94

Nor has only the law changed; the world has too. And when “far-reaching systemic and structural changes” make an “earlier error all the more egregious and harmful,” stare decisis can lose its force.95 In the era when the separate sovereigns exception first emerged, the federal criminal code was new, thin, modest, and restrained. Today, it can make none of those of boasts. Some suggest that “the federal government has [now] duplicated vir-tually every major state crime.”96 Others estimate that the U. S. Code contains more than 4,500 criminal statutes, not even counting the hundreds of thousands of federal regulations that can trigger criminal penalties.97 Still others suggest that ” ‘[t]here is no one in the United States over the age of 18 who cannot be indicted for some federal crime.’ “98 If long ago the Court could have thought “the benignant spirit” of prosecutors rather than unwavering enforcement of the Constitution sufficient protection against the threat of double prosecutions, it’s unclear how we still might.

That leaves reliance. But the only people who have relied on the separate sovereigns exception are prosecutors who have sought to double-prosecute and double-punish. And this Court has long rejected the idea that “law enforcement reliance interests outweig[h] the interest in protecting individual constitutional rights so as to warrant fidelity to an unjustifiable rule.”99 Instead, “[i]f it is clear that a practice is unlawful, individuals’ interest in its discontinuance clearly outweighs any law enforcement ‘entitlement’ to its persistence.”100 That is the case here.

The Court today disregards these lessons. It worries that overturning the separate sovereigns rule could undermine the reliance interests of prosecutors in transnational cases who might be prohibited from trying individuals already acquitted by a foreign court. Ante, at 7. Yet even on its own terms, this argument is unpersuasive. The government has not even attempted to quantify the scope of the alleged “problem,” and perhaps for good reason. Domestic prosecutors regularly coordinate with their foreign counterparts when pursuing transnational criminals, so they can often choose the most favorable forum for their mutual efforts. And because Blockburger requires an identity of elements before the double jeopardy bar can take hold, domestic prosecutors, armed with their own abundant criminal codes, will often be able to find new offenses to charge if they are unsatisfied with outcomes elsewhere.

*  *  *

     Enforcing the Constitution always bears its costs. But when the people adopted the Constitution and its Bill of Rights, they thought the liberties promised there worth the costs. It is not for this Court to reassess this judgment to make the prosecutor’s job easier. Nor is there any doubt that the benefits the framers saw in prohibiting double prosecutions remain real, and maybe more vital than ever, today. When governments may unleash all their might in multiple prosecutions against an individual, exhausting themselves only when those who hold the reins of power are content with the result, it is “the poor and the weak,”101 and the unpopular and controversial, who suffer first–and there is nothing to stop them from being the last. The separate sovereigns exception was wrong when it was invented, and it remains wrong today.