Apprendi v. New Jersey/Concurrence Thomas

by Clarence Thomas

JUSTICE THOMAS, with whom JUSTICE SCALIA joins as to Parts I and II, concurring.

I join the opinion of the Court in full. I write separately to explain my view that the Constitution requires a broader rule than the Court adopts.

I

This case turns on the seemingly simple question of what constitutes a "crime." Under the Federal Constitution, "the accused" has the right (1) "to be informed of the nature and cause of the accusation" (that is, the basis on which he is accused of a crime), (2) to be "held to answer for a capital, or otherwise infamous crime" only on an indictment or presentment of a grand jury, and (3) to be tried by "an impartial jury of the State and district wherein the crime shall have been committed." Amdts. 5 and 6. See also Art. III, §2, cl. 3 ("The Trial of all Crimes . . . shall be by Jury"). With the exception of the Grand Jury Clause, see Hurtado v. California, 110 U.S. 516, 538 (1884), the Court has held that these protections apply in state prosecutions, Herring v. New York, 422 U.S. 853, 857, and n.7 (1975). Further, the Court has held that due process requires that the jury find [p500] beyond a reasonable doubt every fact necessary to constitute the crime. In re Winship, 397 U.S. 358, 364 (1970).

All of these constitutional protections turn on determining which facts constitute the "crime"—that is, which facts are the "elements" or "ingredients" of a crime. In order for an accusation of a crime (whether by indictment or some other form) to be proper under the common law, and thus proper under the codification of the common-law rights in the Fifth and Sixth Amendments, it must allege all elements of that crime; likewise, in order for a jury trial of a crime to be proper, all elements of the crime must be proved to the jury (and, under Winship, proved beyond a reasonable doubt). See J. Story, Commentaries on the Constitution §§ 928–929, pp. 660–662, § 934, p. 664 (1833); J. Archbold, Pleading and Evidence in Criminal Cases *41, *99–*100 (5th Am. ed. 1846) (hereinafter Archbold).^[1]

Thus, it is critical to know which facts are elements. This question became more complicated following the Court's decision in McMillan v. Pennsylvania, 477 U.S. 79 (1986), which spawned a special sort of fact known as a sentencing enhancement. See ante, at 478, 485, 494. Such a fact increases a defendant's punishment but is not subject to the constitutional protections to which elements are subject. JUSTICE O'CONNOR's dissent, in agreement with McMillan and Almendarez-Torres v. United States, 523 U.S. 224 (1998), takes the view that a legislature is free (within unspecified outer limits) to decree which facts are elements and which are sentencing enhancements. Post, at 524.

Sentencing enhancements may be new creatures, but the question that they create for courts is not. Courts have [p501] long had to consider which facts are elements in order to determine the sufficiency of an accusation (usually an indictment). The answer that courts have provided regarding the accusation tells us what an element is, and it is then a simple matter to apply that answer to whatever constitutional right may be at issue in a case—here, Winship and the right to trial by jury. A long line of essentially uniform authority addressing accusations, and stretching from the earliest reported cases after the founding until well into the 20th century, establishes that the original understanding of which facts are elements was even broader than the rule that the Court adopts today.

This authority establishes that a "crime" includes every fact that is by law a basis for imposing or increasing punishment (in contrast with a fact that mitigates punishment). Thus, if the legislature defines some core crime and then provides for increasing the punishment of that crime upon a finding of some aggravating fact—of whatever sort, including the fact of a prior conviction—the core crime and the aggravating fact together constitute an aggravated crime, just as much as grand larceny is an aggravated form of petit larceny. The aggravating fact is an element of the aggravated crime. Similarly, if the legislature, rather than creating grades of crimes, has provided for setting the punishment of a crime based on some fact—such as a fine that is proportional to the value of stolen goods—that fact is also an element. No multi-factor parsing of statutes, of the sort that we have attempted since McMillan, is necessary. One need only look to the kind, degree, or range of punishment to which the prosecution is by law entitled for a given set of facts. Each fact necessary for that entitlement is an element.

II

A

Cases from the founding to roughly the end of the Civil War establish the rule that I have described, applying it to [p502] all sorts of facts, including recidivism. As legislatures varied common-law crimes and created new crimes, American courts, particularly from the 1840's on, readily applied to these new laws the common-law understanding that a fact that is by law the basis for imposing or increasing punishment is an element.^[2]

Massachusetts, which produced the leading cases in the antebellum years, applied this rule as early as 1804, in Commonwealth v. Smith, 1 Mass. *245, and foreshadowed the fuller discussion that was to come. Smith was indicted for and found guilty of larceny, but the indictment failed to allege the value of all of the stolen goods. Massachusetts had abolished the common-law distinction between grand and simple larceny, replacing it with a single offense of larceny whose punishment (triple damages) was based on the value of the stolen goods. The prosecutor relied on this abolition of the traditional distinction to justify the indictment's omissions. The court, however, held that it could not sentence the defendant for the stolen goods whose value was not set out in the indictment. Id., at *246–*247.

The understanding implicit in Smith was explained in Hope v. Commonwealth, 50 Mass. 134 (1845). Hope was indicted for and convicted of larceny. The larceny statute at [p503] issue retained the single-offense structure of the statute addressed in Smith, and established two levels of sentencing based on whether the value of the stolen property exceeded $100. The statute was structured similarly to the statutes that we addressed in Jones v. United States, 526 U.S. 227, 230 (1999), and, even more, Castillo v. United States, ante, at 122, in that it first set out the core crime and then, in subsequent clauses, set out the ranges of punishments.^[3] Further, the statute opened by referring simply to "the offence of larceny," suggesting, at least from the perspective of our post-McMillan cases, that larceny was the crime whereas the value of the stolen property was merely a fact for sentencing. But the matter was quite simple for the Massachusetts high court. Value was an element because punishment varied with value:

"Our statutes, it will be remembered, prescribe the punishment for larceny, with reference to the value of the property stolen; and for this reason, as well as because it is in conformity with long established practice, the court are of opinion that the value of the property alleged to be stolen must be set forth in the indictment." 50 Mass., at 137.

Two years after Hope, the court elaborated on this rule in a case involving burglary, stating that if "certain acts are, by force of the statutes, made punishable with greater severity, when accompanied with aggravating circumstances," then [p504] the statute has "creat[ed] two grades of crime." Larned v. Commonwealth, 53 Mass. 240, 242 (1847). See also id., at 241 ("[T]here is a gradation of offences of the same species" where the statute sets out "various degrees of punishment").

Conversely, where a fact was not the basis for punishment, that fact was, for that reason, not an element. Thus, in Commonwealth v. McDonald, 59 Mass. 365 (1850), which involved an indictment for attempted larceny from the person, the court saw no error in the failure of the indictment to allege any value of the goods that the defendant had attempted to steal. The defendant, in challenging the indictment, apparently relied on Smith and Hope, and the court rejected his challenge by explaining that "[a]s the punishment . . . does not depend on the amount stolen, there was no occasion for any allegation as to value in this indictment." 59 Mass., at 367. See Commonwealth v. Burke, 94 Mass. 182, 183 (1866) (applying same reasoning to completed larceny from the person; finding no trial error where value was not proved to jury).

Similar reasoning was employed by the Wisconsin Supreme Court in Lacy v. State, 15 Wis. *13 (1862), in interpreting a statute that was also similar to the statutes at issue in Jones and Castillo. The statute, in a single paragraph, outlawed arson of a dwelling house at night. Arson that killed someone was punishable by life in prison; arson that did not kill anyone was punishable by 7 to 14 years in prison; arson of a house in which no person was lawfully dwelling was punishable by 3 to 10 years.^[4] The court had no trouble [p505] concluding that the statute "creates three distinct statutory offenses," 15 Wis., at *15, and that the lawful presence of a person in the dwelling was an element of the middle offense. The court reasoned from the gradations of punishment: "That the legislature considered the circumstance that a person was lawfully in the dwelling house when fire was set to it most material and important, and as greatly aggravating the crime, is clear from the severity of the punishment imposed." Id., at *16. The "aggravating circumstances" created "the higher statutory offense[s]." Id., at *17. Because the indictment did not allege that anyone had been present in the dwelling, the court reversed the defendant's 14-year sentence, but, relying on Larned, supra, the court remanded to permit sentencing under the lowest grade of the crime (which was properly alleged in the indictment). 15 Wis., at *17.

Numerous other state and federal courts in this period took the same approach to determining which facts are elements of a crime. See Ritchey v. State, 7 Blackf. 168, 169 (Ind. 1844) (citing Commonwealth v. Smith, 1 Mass. *245 (1804), and holding that indictment for arson must allege value of property destroyed, because statute set punishment based on value); Spencer v. State, 13 Ohio 401, 406, 408 (1844) (holding that value of goods intended to be stolen is not "an ingredient of the crime" of burglary with intent to steal, because punishment under statute did not depend on value; contrasting larceny, in which "[v]alue must be laid, and value proved, that the jury may find it, and the court, by that means, know whether it is grand or petit, and apply the grade of punishment the statute awards"); United States v. Fisher, 25 F. Cas. 1086 (CC Ohio 1849) (McLean, J.) ("A car- [p506] rier of the mail is subject to a higher penalty where he steals a letter out of the mail, which contains an article of value. And when this offense is committed, the indictment must allege the letter contained an article of value, which aggravates the offense and incurs a higher penalty"); Brightwell v. State, 41 Ga. 482, 483 (1871) ("When the law prescribes a different punishment for different phases of the same crime, there is good reason for requiring the indictment to specify which of the phases the prisoner is charged with. The record ought to show that the defendant is convicted of the offense for which he is sentenced"). Cf. State v. Farr, 12 Rich. 24, 29 (S.C. App. 1859) (where two statutes barred purchasing corn from a slave, and one referred to purchasing from slave who lacked a permit, absence of permit was not an element, because both statutes had the same punishment).

Also demonstrating the common-law approach to determining elements was the well-established rule that, if a statute increased the punishment of a common-law crime, whether felony or misdemeanor, based on some fact, then that fact must be charged in the indictment in order for the court to impose the increased punishment. Archbold *106; see id., at *50; ante, at 480–481. There was no question of treating the statutory aggravating fact as merely a sentencing enhancement—as a nonelement enhancing the sentence of the common-law crime. The aggravating fact was an element of a new, aggravated grade of the common-law crime simply because it increased the punishment of the common-law crime. And the common-law crime was, in relation to the statutory one, essentially just like any other lesser included offense. See Archbold *106.

Further evidence of the rule that a crime includes every fact that is by law a basis for imposing or increasing punishment comes from early cases addressing recidivism statutes. As JUSTICE SCALIA has explained, there was a tradition of treating recidivism as an element. See Almendarez-Torres, 523 U.S., at 256–257, 261 (dissenting opinion). That tradi- [p507] tion stretches back to the earliest years of the Republic. See, e.g., Commonwealth v. Welsh, 4 Va. 57 (1817); Smith v. Commonwealth, 14 Serg. & Rawle 69 (Pa. 1826); see also Archbold *695–*696. For my purposes, however, what is noteworthy is not so much the fact of that tradition as the reason for it: Courts treated the fact of a prior conviction just as any other fact that increased the punishment by law. By the same reasoning that the courts employed in Hope, Lacy, and the other cases discussed above, the fact of a prior conviction was an element, together with the facts constituting the core crime of which the defendant was charged, of a new, aggravated crime.

The two leading antebellum cases on whether recidivism is an element were Plumbly v. Commonwealth, 43 Mass. 413 (1841), and Tuttle v. Commonwealth, 68 Mass. 505 (1854). In the latter, the court explained the reason for treating as an element the fact of the prior conviction:

"When the statute imposes a higher penalty upon a second and third conviction, respectively, it makes the prior conviction of a similar offence a part of the description and character of the offence intended to be punished; and therefore the fact of such prior conviction must be charged, as well as proved. It is essential to an indictment, that the facts constituting the offence intended to be punished should be averred." Id., at 506.

The court rested this rule on the common law and the Massachusetts equivalent of the Sixth Amendment's Notice Clause. Ibid. See also Commonwealth v. Haynes, 107 Mass. 194, 198 (1871) (reversing sentence, upon confession of error by attorney general, in case similar to Tuttle).

Numerous other cases treating the fact of a prior conviction as an element of a crime take the same view. They make clear, by both their holdings and their language, that when a statute increases punishment for some core crime based on the fact of a prior conviction, the core crime and [p508] the fact of the prior crime together create a new, aggravated crime. Kilbourn v. State, 9 Conn. 560, 563 (1833) ("No person ought to be, or can be, subjected to a cumulative penalty, without being charged with a cumulative offence"); Plumbly, supra, at 414 (conviction under recidivism statute is "one conviction, upon one aggregate offence"); Hines v. State, 26 Ga. 614, 616 (1859) (reversing enhanced sentence imposed by trial judge and explaining, "[T]he question, whether the offence was a second one, or not, was a question for the jury. . . . The allegation [of a prior offence] is certainly one of the first importance to the accused, for if it is true, he becomes subject to a greatly increased punishment"). See also Commonwealth v. Phillips, 28 Mass. 28, 33 (1831) ("[U]pon a third conviction, the court may sentence the convict to hard labor for life. The punishment is to be awarded upon that conviction, and for the offence of which he is then and there convicted").

Even the exception to this practice of including the fact of a prior conviction in the indictment and trying it to the jury helps to prove the rule that that fact is an element because it increases the punishment by law. In State v. Freeman, 27 Vt. 523 (1855), the Vermont Supreme Court upheld a statute providing that, in an indictment or complaint for violation of a liquor law, it was not necessary to allege a prior conviction of that law in order to secure an increased sentence. But the court did not hold that the prior conviction was not an element; instead, it held that the liquor law created only minor offenses that did not qualify as crimes. Thus, the state constitutional protections that would attach were a "crime" at issue did not apply. Id., at 527; see Goeller v. State, 119 Md. 61, 66–67, 85 A. 954, 956 (1912) (discussing Freeman). At the same time, the court freely acknowledged that it had "no doubt" of the general rule, particularly as articulated in Massachusetts, that "it is necessary to allege the former conviction, in the indictment, when a higher [p509] sentence is claimed on that account." Freeman, supra, at 526. Unsurprisingly, then, a leading treatise explained Freeman as only "apparently" contrary to the general rule and as involving a "special statute." 3 F. Wharton, Criminal Law §3417, p. 307, n. r (7th rev. ed. 1874) (hereinafter Wharton). In addition, less than a decade after Freeman, the same Vermont court held that if a defendant charged with a successive violation of the liquor laws contested identity—that is, whether the person in the record of the prior conviction was the same as the defendant—he should be permitted to have a jury resolve the question. State v. Haynes, 35 Vt. 570, 572–573 (1863). (Freeman itself had anticipated this holding by suggesting the use of a jury to resolve disputes over identity. See 27 Vt., at 528.) In so holding, Haynes all but applied the general rule, since a determination of identity was usually the chief factual issue whenever recidivism was charged. See Archbold *695–*696; see also, e.g., Graham v. West Virginia, 224 U.S. 616, 620–621 (1912) (defendant had been convicted under three different names).^[5] [p510]

B

An 1872 treatise by one of the leading authorities of the era in criminal law and procedure confirms the common-law understanding that the above cases demonstrate. The treatise condensed the traditional understanding regarding the indictment, and thus regarding the elements of a crime, to the following: "The indictment must allege whatever is in law essential to the punishment sought to be inflicted." 1 J. Bishop, Law of Criminal Procedure 50 (2d ed. 1872) (hereinafter Bishop, Criminal Procedure). See id., § 81, at 51 ("[T]he indictment must contain an allegation of every fact which is legally essential to the punishment to be inflicted"); id., § 540, at 330 ("[T]he indictment must . . . contain an averment of every particular thing which enters into the punishment"). Crimes, he explained, consist of those "acts to which the law affixes . . . punishment," id., § 80, at 51, or, stated differently, a crime consists of the whole of "the wrong upon which the punishment is based," id., § 84, at 53. In a later edition, Bishop similarly defined the elements of a crime as "that wrongful aggregation out of which the punishment proceeds." 1 J. Bishop, New Criminal Procedure § 84, p. 49 (4th ed. 1895).

Bishop grounded his definition in both a generalization from well-established common-law practice, 1 Bishop, Criminal Procedure §§ 81–84, at 51–53, and in the provisions of Federal and State Constitutions guaranteeing notice of an accusation in all criminal cases, indictment by a grand jury for serious crimes, and trial by jury. With regard to the common law, he explained that his rule was "not made apparent to our understandings by a single case only, but by all the cases," id., § 81, at 51, and was followed "in all cases, without one exception," id., § 84, at 53. To illustrate, he observed that there are

"various statutes whereby, when . . . assault is committed with a particular [p511] intent, or with a particular weapon, or the like, it is subjected to a particular corresponding punishment, heavier than that for common assault, or differing from it, pointed out by the statute. And the reader will notice that, in all cases where the peculiar or aggravated punishment is to be inflicted, the peculiar or aggravating matter is required to be set out in the indictment." Id., §82, at 52.

He also found burglary statutes illustrative in the same way. Id., § 83, at 52–53. Bishop made no exception for the fact of a prior conviction—he simply treated it just as any other aggravating fact: "[If] it is sought to make the sentence heavier by reason of its being [a second or third offence], the fact thus relied on must be averred in the indictment; because the rules of criminal procedure require the indictment, in all cases, to contain an averment of every fact essential to the punishment sought to be inflicted." 1 J. Bishop, Commentaries on Criminal Law § 961, pp. 564–565 (5th ed. 1872).

The constitutional provisions provided further support, in his view, because of the requirements for a proper accusation at common law and because of the common-law understanding that a proper jury trial required a proper accusation: "The idea of a jury trial, as it has always been known where the common law prevails, includes the allegation, as part of the machinery of the trial . . . . [A]n accusation which lacks any particular fact which the law makes essential to the punishment is . . . no accusation within the requirements of the common law, and it is no accusation in reason." 1 Bishop, Criminal Procedure § 87, at 55. See id., § 88, at 56 (notice and indictment requirements ensure that before "persons held for crimes . . . shall be convicted, there shall be an allegation made against them of every element of crime which the law makes essential to the punishment to be inflicted").

Numerous high courts contemporaneously and explicitly agreed that Bishop had accurately captured the common-law understanding of what facts are elements of a crime. See, [p512] e.g., Hobbs v. State, 44 Tex. 353, 354 (1875) (favorably quoting 1 Bishop, Criminal Procedure § 81); Maguire v. State, 47 Md. 485, 497 (1878) (approvingly citing different Bishop treatise for the same rule); Larney v. Cleveland, 34 Ohio St. 599, 600 (1878) (rule and reason for rule "are well stated by Mr. Bishop"); State v. Hayward, 83 Mo. 299, 307 (1884) (extensively quoting § 81 of Bishop's "admirable treatise"); Riggs v. State, 104 Ind. 261, 262, 3 N.E. 886, 887 (1885) ("We agree with Mr. Bishop that the nature and cause of the accusation are not stated where there is no mention of the full act or series of acts for which the punishment is to be inflicted" (internal quotation marks omitted)); State v. Perley, 86 Me. 427, 431, 30 A. 74, 75 (1894) ("The doctrine of the court, says Mr. Bishop, is identical with that of reason, viz: that the indictment must contain an allegation of every fact which is legally essential to the punishment to be inflicted" (internal quotation marks omitted)); see also United States v. Reese, 92 U.S. 214, 232–233 (1876) (Clifford, J., concurring in judgment) (citing and paraphrasing 1 Bishop, Criminal Procedure § 81).

C

In the half century following publication of Bishop's treatise, numerous courts applied his statement of the common-law understanding; most of them explicitly relied on his treatise. Just as in the earlier period, every fact that was by law a basis for imposing or increasing punishment (including the fact of a prior conviction) was an element. Each such fact had to be included in the accusation of the crime and proved to the jury.

Courts confronted statutes quite similar to the ones with which we have struggled since McMillan, and, applying the traditional rule, they found it not at all difficult to determine whether a fact was an element. In Hobbs, supra, the defendant was indicted for a form of burglary punishable by 2 to 5 years in prison. A separate statutory section provided for an increased sentence, up to double the punishment [p513] to which the defendant would otherwise be subject, if the entry into the house was effected by force exceeding that incidental to burglary. The trial court instructed the jury to sentence the defendant to 2 to 10 years if it found the requisite level of force, and the jury sentenced him to 3. The Texas Supreme Court, relying on Bishop, reversed because the indictment had not alleged such force; even though the jury had sentenced Hobbs within the range (2 to 5 years) that was permissible under the lesser crime that the indictment had charged, the court thought it "impossible to say . . . that the erroneous charge of the court may not have had some weight in leading the jury" to impose the sentence that it did. 44 Tex., at 355.^[6] See also Searcy v. State, 1 Tex. App. 440, 444 (1876) (similar); Garcia v. State, 19 Tex. App. 389, 393 (1885) (not citing Hobbs, but relying on Bishop to reverse 10-year sentence for assault with a bowie-knife or dagger, where statute doubled range for assault from 2 to 7 to 4 to 14 years if the assault was committed with either weapon but where indictment had not so alleged).

As in earlier cases, such as McDonald (discussed supra, at 504), courts also used the converse of the Bishop rule to explain when a fact was not an element of the crime. In Perley, supra, the defendant was indicted for and convicted of robbery, which was punishable by imprisonment for life [p514] or any term of years. The court, relying on Bishop, Hope, McDonald, and other authority, rejected his argument that Maine's Notice Clause (which of course required all elements to be alleged) required the indictment to allege the value of the goods stolen, because the punishment did not turn on value: "[T]here is no provision of this statute which makes the amount of property taken an essential element of the offense; and there is no statute in this State which creates degrees in robbery, or in any way makes the punishment of the offense dependent upon the value of the property taken." 86 Me., at 432, 30 A., at 75. The court further explained that "where the value is not essential to the punishment it need not be distinctly alleged or proved." Id., at 433, 30 A., at 76.

Reasoning similar to Perley and the Texas cases is evident in other cases as well. See Jones v. State, 63 Ga. 141, 143 (1879) (where punishment for burglary in the day is 3 to 5 years in prison and for burglary at night is 5 to 20, time of burglary is a "constituent of the offense"; indictment should "charge all that is requisite to render plain and certain every constituent of the offense"); United States v. Woodruff, 68 F. 536, 538 (Kan. 1895) (where embezzlement statute "contemplates that there should be an ascertainment of the exact sum for which a fine may be imposed" and jury did not determine amount, judge lacked authority to impose fine; "[o]n such an issue the defendant is entitled to his constitutional right of trial by jury").

Courts also, again just as in the pre-Bishop period, applied the same reasoning to the fact of a prior conviction as they did to any other fact that aggravated the punishment by law. Many, though far from all, of these courts relied on Bishop. In 1878, Maryland's high court, in Maguire v. State, 47 Md. 485, stated the rule and the reason for it in language indistinguishable from that of Tuttle a quarter century before:

"The law would seem to be well settled, that if the party be proceeded against for a second or third offence [p515] under the statute, and the sentence prescribed be different from the first, or severer, by reason of its being such second or third offence, the fact thus relied on must be averred in the indictment; for the settled rule is, that the indictment must contain an averment of every fact essential to justify the punishment inflicted." Maguire, supra, at 496 (citing English cases, Plumbly v. Commonwealth, 43 Mass. 413 (1841), Wharton, and Bishop).

In Goeller v. State, 119 Md. 61, 85 A. 954 (1912), the same court reaffirmed Maguire and voided, as contrary to Maryland's Notice Clause, a statute that permitted the trial judge to determine the fact of a prior conviction. The court extensively quoted Bishop, who had, in the court's view, treated the subject "more fully, perhaps, than any other legal writer," and it cited, among other authorities, "a line of Massachusetts decisions" and Riggs (quoted supra, at 14). 119 Md., at 66, 85 A., at 955. In Larney, 34 Ohio St., at 600–601, the Supreme Court of Ohio, in an opinion citing only Bishop, reversed a conviction under a recidivism statute where the indictment had not alleged any prior conviction. (The defendant had also relied on Plumbly, supra, and Kilbourn v. State, 9 Conn. 560 (1833). 34 Ohio St., at 600.) And in State v. Adams, 64 N.H. 440, 13 A. 785 (1888), the court, relying on Bishop, explained that "[t]he former conviction being a part of the description and character of the offense intended to be punished, because of the higher penalty imposed, it must be alleged." Id., at 442, 13 A., at 786. The defendant had been "charged with an offense aggravated by its repetitious character." Ibid. See also Evans v. State, 150 Ind. 651, 653, 50 N.E. 820 (1898) (similar); Shiflett v. Commonwealth, 114 Va. 876, 877, 77 S.E. 606, 607 (1913) (similar).

Even without any reliance on Bishop, other courts addressing recidivism statutes employed the same reasoning as did he and the above cases—that a crime includes any fact to which punishment attaches. One of the leading cases was [p516] Wood v. People, 53 N.Y. 511 (1873). The statute in Wood provided for increased punishment if the defendant had previously been convicted of a felony then discharged from the conviction. The court, repeatedly referring to "the aggravated offence," id., at 513, 515, held that the facts of the prior conviction and of the discharge must be proved to the jury, for "[b]oth enter into and make a part of the offence . . . subjecting the prisoner to the increased punishment." Id., at 513; see ibid. (fact of prior conviction was an "essential ingredient" of the offense). See also Johnson v. People, 55 N.Y. 512, 514 (1874) ("A more severe penalty is denounced by the statute for a second offence; and all the facts to bring the case within the statute must be [alleged in the indictment and] established on the trial"); People v. Sickles, 156 N.Y. 541, 544–545, 51 N.E. 288, 289 (1898) (reaffirming Wood and Johnson and explaining that "the charge is not merely that the prisoner has committed the offense specifically described, but that, as a former convict, his second offense has subjected him to an enhanced penalty").

Contemporaneously with the New York Court of Appeals in Wood and Johnson, state high courts in California and Pennsylvania offered similar explanations for why the fact of a prior conviction is an element. In People v. Delany, 49 Cal. 394 (1874), which involved a statute making petit larceny (normally a misdemeanor) a felony if committed following a prior conviction for petit larceny, the court left no doubt that the fact of the prior conviction was an element of an aggravated crime consisting of petit larceny committed following a prior conviction for petit larceny:

"The particular circumstances of the offense are stated [in the indictment], and consist of the prior convictions and of the facts constituting the last larceny.

·

·

·

·

·

"[T]he former convictions are made to adhere to and constitute a portion of the aggravated offense." Id., at 395.

[p517] "The felony consists both of the former convictions and of the particular larceny. . . . [T]he former convictions were a separate fact; which, taken in connection with the facts constituting the last offense, make a distinct and greater offense than that charged, exclusive of the prior convictions." Id., at 396.^[7]

See also People v. Coleman, 145 Cal. 609, 610–611, 79 P. 283, 284–285 (1904).

Similarly, in Rauch v. Commonwealth, 78 Pa. 490 (1876), the court applied its 1826 decision in Smith v. Commonwealth, 14 Serg. & Rawle 69, and reversed the trial court's imposition of an enhanced sentence "upon its own knowledge of its records." 78 Pa., at 494. The court explained that "imprisonment in jail is not a lawful consequence of a mere conviction for an unlawful sale of liquors. It is the lawful consequence of a second sale only after a former conviction. On every principle of personal security and the due administration of justice, the fact which gives rightfulness to the greater punishment should appear in the record." Ibid. See also id., at 495 ("But clearly the substantive offence, which draws to itself the greater punishment, is the unlawful sale after a former conviction. This, therefore, is the very offence he is called upon to defend against").

Meanwhile, Massachusetts reaffirmed its earlier decisions, striking down, in Commonwealth v. Harrington, 130 Mass. 35 (1880), a liquor law that provided a small fine for a first or second conviction, provided a larger fine or imprisonment up to a year for a third conviction, and specifically provided that a prior conviction need not be alleged in the complaint. The court found this law plainly inconsistent with Tuttle and with the State's Notice Clause, explaining that "the offence which is punishable with the higher penalty is not fully and [p518] substantially described to the defendant, if the complaint fails to set forth the former convictions which are essential features of it." 130 Mass., at 36.^[8]

Without belaboring the point any further, I simply note that this traditional understanding—that a "crime" includes every fact that is by law a basis for imposing or increasing punishment—continued well into the 20th century, at least until the middle of the century. See Knoll & Singer, Searching for the "Tail of the Dog": Finding "Elements" of Crimes in the Wake of McMillan v. Pennsylvania, 22 Seattle U. L. Rev. 1057, 1069–1081 (1999) (surveying 20th century decisions of federal courts prior to McMillan); see also People v. Ratner, 67 Cal. App. 2d Supp. 902, 903–906, 153 P.2d 790, 791–793 (1944). In fact, it is fair to say that McMillan began a revolution in the law regarding the definition of "crime." Today's decision, far from being a sharp break with the past, marks nothing more than a return to the status quo ante—the status quo that reflected the original meaning of the Fifth and Sixth Amendments.

III

The consequence of the above discussion for our decisions in Almendarez-Torres and McMillan should be plain enough, but a few points merit special mention.

[p519] First, it is irrelevant to the question of which facts are elements that legislatures have allowed sentencing judges discretion in determining punishment (often within extremely broad ranges). See ante, at 481–482; post, at 544–545 (O'CONNOR, J., dissenting). Bishop, immediately after setting out the traditional rule on elements, explained why:

"The reader should distinguish between the foregoing doctrine, and the doctrine . . . that, within the limits of any discretion as to the punishment which the law may have allowed, the judge, when he pronounces sentence, may suffer his discretion to be influenced by matter shown in aggravation or mitigation, not covered by the allegations of the indictment. . . . The aggravating circumstances spoken of cannot swell the penalty above what the law has provided for the acts charged against the prisoner, and they are interposed merely to check the judicial discretion in the exercise of the permitted mercy [in finding mitigating circumstances]. This is an entirely different thing from punishing one for what is not alleged against him." 1 Bishop, Criminal Procedure §85, at 54.

See also 1 J. Bishop, New Commentaries on the Criminal Law §§ 600–601, pp. 370–371, § 948, p. 572 (8th ed. 1892) (similar). In other words, establishing what punishment is available by law and setting a specific punishment within the bounds that the law has prescribed are two different things.^[9] [p520] Cf. 4 W. Blackstone, Commentaries on the Law of England 371–372 (1769) (noting judges' broad discretion in setting amount of fine and length of imprisonment for misdemeanors, but praising determinate punishment and "discretion . . . regulated by law"); Perley, 86 Me., at 429, 432, 30 A., at 74, 75–76 (favorably discussing Bishop's rule on elements without mentioning, aside from quotation of statute in statement of facts, that defendant's conviction for robbery exposed him to imprisonment for life or any term of years). Thus, it is one thing to consider what the Constitution requires the prosecution to do in order to entitle itself to a particular kind, degree, or range of punishment of the accused, see Woodruff, 68 F., at 538, and quite another to consider what constitutional constraints apply either to the imposition of punishment within the limits of that entitlement or to a legislature's ability to set broad ranges of punishment. In answering the former constitutional question, I need not, and do not, address the latter.

Second, and related, one of the chief errors of Almendarez-Torres—an error to which I succumbed—was to attempt to discern whether a particular fact is traditionally (or typically) a basis for a sentencing court to increase an offender's sentence. 523 U.S., at 243–244; see id., at 230, 241. For the [p521] reasons I have given, it should be clear that this approach just defines away the real issue. What matters is the way by which a fact enters into the sentence. If a fact is by law the basis for imposing or increasing punishment—for establishing or increasing the prosecution's entitlement—it is an element. (To put the point differently, I am aware of no historical basis for treating as a nonelement a fact that by law sets or increases punishment.) When one considers the question from this perspective, it is evident why the fact of a prior conviction is an element under a recidivism statute. Indeed, cases addressing such statutes provide some of the best discussions of what constitutes an element of a crime. One reason frequently offered for treating recidivism differently, a reason on which we relied in Almendarez-Torres, supra, at 235, is a concern for prejudicing the jury by informing it of the prior conviction. But this concern, of which earlier courts were well aware, does not make the traditional understanding of what an element is any less applicable to the fact of a prior conviction. See, e.g., Maguire, 47 Md., at 498; Sickles, 156 N.Y., at 547, 51 N.E., at 290.^[10]

Third, I think it clear that the common-law rule would cover the McMillan situation of a mandatory minimum sentence (in that case, for visible possession of a firearm during the commission of certain crimes). No doubt a defendant could, under such a scheme, find himself sentenced to the same term to which he could have been sentenced absent the mandatory minimum. The range for his underlying crime [p522] could be 0 to 10 years, with the mandatory minimum of 5 years, and he could be sentenced to 7. (Of course, a similar scenario is possible with an increased maximum.) But it is equally true that his expected punishment has increased as a result of the narrowed range and that the prosecution is empowered, by invoking the mandatory minimum, to require the judge to impose a higher punishment than he might wish. The mandatory minimum "entitl[es] the government," Woodruff, supra, at 538, to more than it would otherwise be entitled (5 to 10 years, rather than 0 to 10 and the risk of a sentence below 5). Thus, the fact triggering the mandatory minimum is part of "the punishment sought to be inflicted," Bishop, Criminal Procedure, at 50; it undoubtedly "enters into the punishment" so as to aggravate it, id., §540, at 330, and is an "ac[t] to which the law affixes . . . punishment," id., § 80, at 51. Further, just as in Hobbs and Searcy, see supra, at 512–513, it is likely that the change in the range available to the judge affects his choice of sentence. Finally, in numerous cases, such as Lacy, Garcia, and Jones, see supra, at 504–505, 514, the aggravating fact raised the whole range—both the top and bottom. Those courts, in holding that such a fact was an element, did not bother with any distinction between changes in the maximum and the minimum. What mattered was simply the overall increase in the punishment provided by law. And in several cases, such as Smith and Woodruff, see supra, at 502, 514, the very concept of maximums and minimums had no applicability, yet the same rule for elements applied. See also Harrington (discussed supra, at 517–518).

Finally, I need not in this case address the implications of the rule that I have stated for the Court's decision in Walton v. Arizona, 497 U.S. 639, 647–649 (1990). See ante, at 496. Walton did approve a scheme by which a judge, rather than a jury, determines an aggravating fact that makes a convict eligible for the death penalty, and thus eligible for a greater punishment. In this sense, that fact is an element. But that scheme exists in a unique context, for in the area of cap- [p523] ital punishment, unlike any other area, we have imposed special constraints on a legislature's ability to determine what facts shall lead to what punishment—we have restricted the legislature's ability to define crimes. Under our recent capital-punishment jurisprudence, neither Arizona nor any other jurisdiction could provide—as, previously, it freely could and did—that a person shall be death eligible automatically upon conviction for certain crimes. We have interposed a barrier between a jury finding of a capital crime and a court's ability to impose capital punishment. Whether this distinction between capital crimes and all others, or some other distinction, is sufficient to put the former outside the rule that I have stated is a question for another day.^[11] *** For the foregoing reasons, as well as those given in the Court's opinion, I agree that the New Jersey procedure at issue is unconstitutional.

Notes

1. JUSTICE O'CONNOR mischaracterizes my argument. See post, at 527–528 (dissenting opinion). Of course the Fifth and Sixth Amendments did not codify common-law procedure wholesale. Rather, and as Story notes, they codified a few particular common-law procedural rights. As I have explained, the scope of those rights turns on what constitutes a "crime." In answering that question, it is entirely proper to look to the common law.
    
2. It is strange that JUSTICE O'CONNOR faults me for beginning my analysis with cases primarily from the 1840's, rather from the time of the founding. See post, at 527–528 (dissenting opinion). As the Court explains, ante, at 478–480, and as she concedes, post, at 525 (O'CONNOR, J., dissenting), the very idea of a sentencing enhancement was foreign to the common law of the time of the founding. JUSTICE O'CONNOR therefore, and understandably, does not contend that any history from the founding supports her position. As far as I have been able to tell, the argument that a fact that was by law the basis for imposing or increasing punishment might not be an element did not seriously arise (at least not in reported cases) until the 1840's. As I explain below, from that time on—for at least a century—essentially all authority rejected that argument, and much of it did so in reliance upon the common law. I find this evidence more than sufficient.
    
3. The Massachusetts statute provided: "Every person who shall commit the offence of larceny, by stealing of the property of another any money, goods or chattels [or other sort of property], if the property stolen shall exceed the value of one hundred dollars, shall be punished by imprisonment in the state prison, not more than five years, or by fine not exceeding six hundred dollars, and imprisonment in the county jail, not more than two years; and if the property stolen shall not exceed the value of one hundred dollars, he shall be punished by imprisonment in the state prison or the county jail, not more than one year, or by fine not exceeding three hundred dollars." Mass. Rev. Stat., ch. 126, § 17 (1836).
    
4. The Wisconsin statute provided: "Every person who shall willfully and maliciously burn, in the night time, the dwelling house of another, whereby the life of any person shall be destroyed, or shall in the night time willfully and maliciously set fire to any other building, owned by himself or another, by the burning whereof such dwelling house shall be burnt in the night time, whereby the life of any person shall be destroyed, shall suffer the same punishment as provided for the crime of murder in the second degree; but if the life of no person shall have been destroyed, he shall be punished by imprisonment in the state prison, not more than fourteen [p505] years nor less than seven years; and if at the time of committing the offense there was no person lawfully in the dwelling house so burnt, he shall be punished by imprisonment in the state prison, not more than ten years nor less than three years." Wis. Rev. Stat., ch. 165, § 1 (1858). The punishment for second-degree murder was life in prison. Ch. 164, § 2.
    
5. Some courts read State v. Smith, 8 Rich. 460 (S.C. App. 1832), a South Carolina case, to hold that the indictment need not allege a prior conviction in order for the defendant to suffer an enhanced punishment. See, e.g., State v. Burgett, 22 Ark. 323, 324 (1860) (so reading Smith and questioning its correctness). The Smith court's holding was somewhat unclear because the court did not state whether the case involved a first or second offense—if a first, the court was undoubtedly correct in rejecting the defendant's challenge to the indictment, because there is no need in an indictment to negate the existence of any prior offense. See Burgett, supra, at 324 (reading indictment that was silent about prior offenses as only charging first offense and as sufficient for that purpose). In addition, the Smith court did not acknowledge the possibility of disputes over identity. Finally, the extent to which the court's apparent holding was followed in practice in South Carolina is unclear, and subsequent South Carolina decisions acknowledged that Smith was out of step with the general rule. See State v. Parris, 89 S.C. 140, 141, 71 S.E. 808, 809 (1911); State v. Mitchell, 220 S.C. 433, 434–436, 68 S.E. 2d 350, 351–352 (1951).
    
6. The gulf between the traditional approach to determining elements and that of our recent cases is manifest when one considers how one might, from the perspective of those cases, analyze the issue in Hobbs. The chapter of the Texas code addressing burglary was entitled simply "Of Burglary" and began with a section explicitly defining "the offense of burglary." After a series of sections defining terms, it then set out six separate sections specifying the punishment for various kinds of burglary. The section regarding force was one of these. See 1 G. Paschal, Digest of Laws of Texas, Part II, Tit. 20, ch. 6, pp. 462–463 (4th ed. 1875). Following an approach similar to that in Almendarez-Torres v. United States, 523 U.S. 224, 231–234, 242–246 (1998), and Castillo v. United States, ante, 124–125, one would likely find a clear legislative intent to make force a sentencing enhancement rather than an element.
    
7. The court held that a general plea of "guilty" to an indictment that includes an allegation of a prior conviction applies to the fact of the prior conviction.
    
8. See also State v. Austin, 113 Mo. 538, 542, 21 S.W. 31, 32 (1893) (prior conviction is a "material fac[t]" of the "aggravated offense"); Bandy v. Hehn, 10 Wyo. 167, 172–174, 67 P. 979, 980 (1902) ("[I]n reason, and by the great weight of authority, as the fact of a former conviction enters into the offense to the extent of aggravating it and increasing the punishment, it must be alleged in the information and proved like any other material fact, if it is sought to impose the greater penalty. The statute makes the prior conviction a part of the description and character of the offense intended to be punished" (citing Tuttle v. Commonwealth, 68 Mass. 505 (1854))); State v. Smith, 129 Iowa 709, 711–712, 106 N.W. 187, 188–189 (1906) (similar); State v. Scheminisky, 31 Idaho 504, 506–507, 174 P. 611, 611–612 (1918) (similar).
    
9. This is not to deny that there may be laws on the borderline of this distinction. In Brightwell v. State, 41 Ga. 482 (1871), the court stated a rule for elements equivalent to Bishop's, then held that whether a defendant had committed arson in the day or at night need not be in the indictment. The court explained that there was "no provision that arson in the night shall be punished for any different period" than arson in the day (both being punishable by 2 to 7 years in prison). Id., at 483. Although there was a statute providing that "arson in the day time shall be punished for a less period than arson in the night time," the court concluded that it merely set "a rule for the exercise of [the sentencing [p520] judge's] discretion" by specifying a particular fact for the judge to consider along with the many others that would enter into his sentencing decision. Ibid. Cf. Jones v. State, 63 Ga. 141, 143 (1879) (whether burglary occurred in day or at night is a "constituent of the offense" because law fixes different ranges of punishment based on this fact). And the statute attached no definite consequence to that particular fact: A sentencing judge presumably could have imposed a sentence of seven years less one second for daytime arson. Finally, it is likely that the statute in Brightwell, given its language ("a less period") and its placement in a separate section, was read as setting out an affirmative defense or mitigating circumstance. See Wright v. State, 113 Ga. App. 436, 437–438, 148 S.E.2d 333, 335–336 (1966) (suggesting that it would be error to refuse to charge later version of this statute to jury upon request of defendant). See generally Archbold *52, *105–*106 (discussing rules for determining whether fact is an element or a defense).
    
10. In addition, it has been common practice to address this concern by permitting the defendant to stipulate to the prior conviction, in which case the charge of the prior conviction is not read to the jury, or, if the defendant decides not to stipulate, to bifurcate the trial, with the jury only considering the prior conviction after it has reached a guilty verdict on the core crime. See, e.g., 1 J. Bishop, Criminal Law § 964, at 566–567 (5th ed. 1872) (favorably discussing English practice of bifurcation); People v. Saunders, 5 Cal. 4th 580, 587–588, 853 P.2d 1093, 1095–1096 (1993) (detailing California approach, since 1874, of permitting stipulation and, more recently, of also permitting bifurcation).
     
11. It is likewise unnecessary to consider whether (and, if so, how) the rule regarding elements applies to the Sentencing Guidelines, given the unique status that they have under Mistretta v. United States, 488 U.S. 361 (1989). But it may be that this special status is irrelevant, because the Guidelines "have the force and effect of laws." Id., at 413 (SCALIA, J., dissenting).