SMITH v. DOE

JUSTICE KENNEDY delivered the opinion of the Court.

The Alaska Sex Offender Registration Act requires convicted sex offenders to register with law enforcement authorities, and much of the information is made public. We must decide whether the registration requirement is a retroactive punishment prohibited by the Ex Post Facto Clause.

I

A

The State of Alaska enacted the Alaska Sex Offender Registration Act (Act) on May 12, 1994. 1994 Alaska Sess. Laws ch. 41. Like its counterparts in other States, the Act is termed a "Megan's Law." Megan Kanka was a 7-year-old New Jersey girl who was sexually assaulted and murdered in 1994 by a neighbor who, unknown to the victim's family, had prior convictions for sex offenses against children. The crime gave impetus to laws for mandatory registration of sex offenders and corresponding community notification. In 1994, Congress passed the Jacob Wetterling Crimes Against Children and Sexually Violent Offender Registration Act, title 17, 108 Stat. 2038, as amended, 42 U. S. C. § 14071, which conditions certain federal law enforcement funding on the States' adoption of sex offender registration laws and sets minimum standards for state programs. By 1996, every State, the District of Columbia, and the Federal Government had enacted some variation of Megan's Law.

The Alaska law, which is our concern in this case, contains two components: a registration requirement and a notification system. Both are retroactive. 1994 Alaska Sess. Laws ch. 41, § 12(a). The Act requires any "sex offender or child kidnapper who is physically present in the state" to register, either with the Department of Corrections (if the individual is incarcerated) or with the local law enforcement authorities (if the individual is at liberty). Alaska Stat. §§ 12.63.010(a), (b) (2000). Prompt registration is mandated. If still in prison, a covered sex offender must register within 30 days before release; otherwise he must do so within a working day of his conviction or of entering the State. § 12.63.010(a). The sex offender must provide his name, aliases, identifying features, address, place of employment, date of birth, conviction information, driver's license number, information about vehicles to which he has access, and postconviction treatment history. § 12.63.010(b)(1). He must permit the authorities to photograph and fingerprint him. § 12.63.010(b)(2).

If the offender was convicted of a single, nonaggravated sex crime, he must provide annual verification of the submitted information for 15 years. §§ 12.63.010(d)(1), 12.63.020(a)(2). If he was convicted of an aggravated sex offense or of two or more sex offenses, he must register for life and verify the information quarterly. §§ 12.63.010(d)(2), 12.63.020(a)(1). The offender must notify his local police department if he moves. § 12.63.010(c). A sex offender who knowingly fails to comply with the Act is subject to criminal prosecution. §§ 11.56.835, 11.56.840.

The information is forwarded to the Alaska Department of Public Safety, which maintains a central registry of sex offenders. § 18.65.087(a). Some of the data, such as fingerprints, driver's license number, anticipated change of address, and whether the offender has had medical treatment afterwards, are kept confidential. §§ 12.63.010(b), 18.65.087(b). The following information is made available to the public: "the sex offender's or child kidnapper's name, aliases, address, photograph, physical description, description[,] license [and] identification numbers of motor vehicles, place of employment, date of birth, crime for which convicted, date of conviction, place and court of conviction, length and conditions of sentence, and a statement as to whether the offender or kidnapper is in compliance with [the update] requirements . . . or cannot be located." § 18.65.087(b). The Act does not specify the means by which the registry information must be made public. Alaska has chosen to make most of the nonconfidential information available on the Internet.

B

Respondents John Doe I and John Doe II were convicted of sexual abuse of a minor, an aggravated sex offense. John Doe I pleaded nolo contendere after a court determination that he had sexually abused his daughter for two years, when she was between the ages of 9 and 11; John Doe II entered a nolo contendere plea to sexual abuse of a 14-year-old child. Both were released from prison in 1990 and completed rehabilitative programs for sex offenders. Although convicted before the passage of the Act, respondents are covered by it. After the initial registration, they are required to submit quarterly verifications and notify the authorities of any changes. Both respondents, along with respondent Jane Doe, wife of John Doe I, brought an action under Rev. Stat. § 1979, 42 U.S.C. § 1983, seeking to declare the Act void as to them under the Ex Post Facto Clause of Article I, § 10, cl. 1, of the Constitution and the Due Process Clause of § 1 of the Fourteenth Amendment. The United States District Court for the District of Alaska granted summary judgment for petitioners. In agreement with the District Court, the Court of Appeals for the Ninth Circuit determined the state legislature had intended the Act to be a nonpunitive, civil regulatory scheme; but, in disagreement with the District Court, it held the effects of the Act were punitive despite the legislature's intent. In consequence, it held the Act violates the Ex Post Facto Clause. Doe I v. Otte, 259 F.3d 979 (2001). We granted certiorari. 534 U.S. 1126 (2002).

II

This is the first time we have considered a claim that a sex offender registration and notification law constitutes retroactive punishment forbidden by the Ex Post Facto Clause. The framework for our inquiry, however, is well established. We must "ascertain whether the legislature meant the statute to establish `civil' proceedings." Kansas v. Hendricks, 521 U.S. 346, 361 (1997). If the intention of the legislature was to impose punishment, that ends the inquiry. If, however, the intention was to enact a regulatory scheme that is civil and nonpunitive, we must further examine whether the statutory scheme is "`so punitive either in purpose or effect as to negate [the State's] intention' to deem it `civil.'" Ibid. (quoting United States v. Ward, 448 U.S. 242, 248-249 (1980)). Because we "ordinarily defer to the legislature's stated intent," Hendricks, supra, at 361, "`only the clearest proof' will suffice to override legislative intent and transform what has been denominated a civil remedy into a criminal penalty," Hudson v. United States, 522 U.S. 93, 100 (1997) (quoting Ward, supra, at 249); see also Hendricks, supra, at 361; United States v. Ursery, 518 U.S. 267, 290 (1996); United States v. One Assortment of 89 Firearms, 465 U.S. 354, 365 (1984).

A

Whether a statutory scheme is civil or criminal "is first of all a question of statutory construction." Hendricks, supra, at 361 (internal quotation marks omitted); see also Hudson, supra, at 99. We consider the statute's text and its structure to determine the legislative objective. Flemming v. Nestor, 363 U.S. 603, 617 (1960). A conclusion that the legislature intended to punish would satisfy an ex post facto challenge without further inquiry into its effects, so considerable deference must be accorded to the intent as the legislature has stated it.

The courts "must first ask whether the legislature, in establishing the penalizing mechanism, indicated either expressly or impliedly a preference for one label or the other." Hudson, supra, at 99 (internal quotation marks omitted). Here, the Alaska Legislature expressed the objective of the law in the statutory text itself. The legislature found that "sex offenders pose a high risk of reoffending," and identified "protecting the public from sex offenders" as the "primary governmental interest" of the law. 1994 Alaska Sess. Laws ch. 41, § 1. The legislature further determined that "release of certain information about sex offenders to public agencies and the general public will assist in protecting the public safety." Ibid. As we observed in Hendricks, where we examined an ex post facto challenge to a postincarceration confinement of sex offenders, an imposition of restrictive measures on sex offenders adjudged to be dangerous is "a legitimate nonpunitive governmental objective and has been historically so regarded." 521 U. S., at 363. In this case, as in Hendricks, "[n]othing on the face of the statute suggests that the legislature sought to create anything other than a civil . . . scheme designed to protect the public from harm." Id., at 361.

Respondents seek to cast doubt upon the nonpunitive nature of the law's declared objective by pointing out that the Alaska Constitution lists the need for protecting the public as one of the purposes of criminal administration. Brief for Respondents 23 (citing Alaska Const., Art. I, § 12). As the Court stated in Flemming v. Nestor, rejecting an ex post facto challenge to a law terminating benefits to deported aliens, where a legislative restriction "is an incident of the State's power to protect the health and safety of its citizens," it will be considered "as evidencing an intent to exercise that regulatory power, and not a purpose to add to the punishment." 363 U. S., at 616 (citing Hawker v. New York, 170 U.S. 189 (1898)). The Court repeated this principle in 89 Firearms, upholding a statute requiring forfeiture of unlicensed firearms against a double jeopardy challenge. The Court observed that, in enacting the provision, Congress "`was concerned with the widespread traffic in firearms and with their general availability to those whose possession thereof was contrary to the public interest.'" 465 U. S., at 364 (quoting Huddleston v. United States, 415 U.S. 814, 824 (1974)). This goal was "plainly more remedial than punitive." 465 U. S., at 364. These precedents instruct us that even if the objective of the Act is consistent with the purposes of the Alaska criminal justice system, the State's pursuit of it in a regulatory scheme does not make the objective punitive.

Other formal attributes of a legislative enactment, such as the manner of its codification or the enforcement procedures it establishes, are probative of the legislature's intent. See Hendricks, supra, at 361; Hudson, supra, at 103; 89 Firearms, supra, at 363. In this case these factors are open to debate. The notification provisions of the Act are codified in the State's "Health, Safety, and Housing Code," § 18, confirming our conclusion that the statute was intended as a nonpunitive regulatory measure. Cf. Hendricks, supra, at 361 (the State's "objective to create a civil proceeding is evidenced by its placement of the Act within the [State's] probate code, instead of the criminal code" (citations omitted)). The Act's registration provisions, however, are codified in the State's criminal procedure code, and so might seem to point in the opposite direction. These factors, though, are not dispositive. The location and labels of a statutory provision do not by themselves transform a civil remedy into a criminal one. In 89 Firearms, the Court held a forfeiture provision to be a civil sanction even though the authorizing statute was in the criminal code. 465 U. S., at 364-365. The Court rejected the argument that the placement demonstrated Congress' "intention to create an additional criminal sanction," observing that "both criminal and civil sanctions may be labeled `penalties.'" Id., at 364, n. 6.

The same rationale applies here. Title 12 of Alaska's Code of Criminal Procedure (where the Act's registration provisions are located) contains many provisions that do not involve criminal punishment, such as civil procedures for disposing of recovered and seized property, Alaska Stat. § 12.36.010 et seq. (2000); laws protecting the confidentiality of victims and witnesses, § 12.61.010 et seq.; laws governing the security and accuracy of criminal justice information, § 12.62.110 et seq.; laws governing civil postconviction actions, § 12.72.010 et seq.; and laws governing actions for writs of habeas corpus, § 12.75.010 et seq., which under Alaska law are "independent civil proceeding[s]," State v. Hannagan, 559 P.2d 1059, 1063 (Alaska 1977). Although some of these provisions relate to criminal administration, they are not in themselves punitive. The partial codification of the Act in the State's criminal procedure code is not sufficient to support a conclusion that the legislative intent was punitive.

The procedural mechanisms to implement the Act do not alter our conclusion. After the Act's adoption Alaska amended its Rules of Criminal Procedure concerning the acceptance of pleas and the entering of criminal judgments. The rule on pleas now requires the court to "infor[m] the defendant in writing of the requirements of [the Act] and, if it can be determined by the court, the period of registration required." Alaska Rule Crim. Proc. 11(c)(4) (2002). Similarly, the written judgments for sex offenses and child kidnapings "must set out the requirements of [the Act] and, if it can be determined by the court, whether that conviction will require the offender or kidnapper to register for life or a lesser period." Alaska Stat. § 12.55.148(a) (2000).

The policy to alert convicted offenders to the civil consequences of their criminal conduct does not render the consequences themselves punitive. When a State sets up a regulatory scheme, it is logical to provide those persons subject to it with clear and unambiguous notice of the requirements and the penalties for noncompliance. The Act requires registration either before the offender's release from confinement or within a day of his conviction (if the offender is not imprisoned). Timely and adequate notice serves to apprise individuals of their responsibilities and to ensure compliance with the regulatory scheme. Notice is important, for the scheme is enforced by criminal penalties. See §§ 11.56.835, 11.56.840. Although other methods of notification may be available, it is effective to make it part of the plea colloquy or the judgment of conviction. Invoking the criminal process in aid of a statutory regime does not render the statutory scheme itself punitive.

Our conclusion is strengthened by the fact that, aside from the duty to register, the statute itself mandates no procedures. Instead, it vests the authority to promulgate implementing regulations with the Alaska Department of Public Safety, §§ 12.63.020(b), 18.65.087(d) — an agency charged with enforcement of both criminal and civil regulatory laws. See, e. g., § 17.30.100 (enforcement of drug laws); § 18.70.010 (fire protection); § 28.05.011 (motor vehicles and road safety); § 44.41.020 (protection of life and property). The Act itself does not require the procedures adopted to contain any safeguards associated with the criminal process. That leads us to infer that the legislature envisioned the Act's implementation to be civil and administrative. By contemplating "distinctly civil procedures," the legislature "indicate[d] clearly that it intended a civil, not a criminal sanction." Ursery, 518 U. S., at 289 (internal quotation marks omitted; alteration in original).

We conclude, as did the District Court and the Court of Appeals, that the intent of the Alaska Legislature was to create a civil, nonpunitive regime.

B

In analyzing the effects of the Act we refer to the seven factors noted in Kennedy v. Mendoza-Martinez, 372 U.S. 144, 168-169 (1963), as a useful framework. These factors, which migrated into our ex post facto case law from double jeopardy jurisprudence, have their earlier origins in cases under the Sixth and Eighth Amendments, as well as the Bill of Attainder and the Ex Post Facto Clauses. See id., at 168-169, and nn. 22-28. Because the Mendoza-Martinez factors are designed to apply in various constitutional contexts, we have said they are "neither exhaustive nor dispositive," United States v. Ward, 448 U. S., at 249; 89 Firearms, 465 U. S., at 365, n. 7, but are "useful guideposts," Hudson, 522 U. S., at 99. The factors most relevant to our analysis are whether, in its necessary operation, the regulatory scheme: has been regarded in our history and traditions as a punishment; imposes an affirmative disability or restraint; promotes the traditional aims of punishment; has a rational connection to a nonpunitive purpose; or is excessive with respect to this purpose.

A historical survey can be useful because a State that decides to punish an individual is likely to select a means deemed punitive in our tradition, so that the public will recognize it as such. The Court of Appeals observed that the sex offender registration and notification statutes "are of fairly recent origin," 259 F. 3d, at 989, which suggests that the statute was not meant as a punitive measure, or, at least, that it did not involve a traditional means of punishing. Respondents argue, however, that the Act — and, in particular, its notification provisions — resemble shaming punishments of the colonial period. Brief for Respondents 33-34 (citing A. Earle, Curious Punishments of Bygone Days 1-2 (1896)).

Some colonial punishments indeed were meant to inflict public disgrace. Humiliated offenders were required "to stand in public with signs cataloguing their offenses." Hirsch, From Pillory to Penitentiary: The Rise of Criminal Incarceration in Early Massachusetts, 80 Mich. L. Rev. 1179, 1226 (1982); see also L. Friedman, Crime and Punishment in American History 38 (1993). At times the labeling would be permanent: A murderer might be branded with an "M," and a thief with a "T." R. Semmes, Crime and Punishment in Early Maryland 35 (1938); see also Massaro, Shame, Culture, and American Criminal Law, 89 Mich. L. Rev. 1880, 1913 (1991). The aim was to make these offenders suffer "permanent stigmas, which in effect cast the person out of the community." Ibid.; see also Friedman, supra, at 40; Hirsch, supra, at 1228. The most serious offenders were banished, after which they could neither return to their original community nor, reputation tarnished, be admitted easily into a new one. T. Blomberg & K. Lucken, American Penology: A History of Control 30-31 (2000). Respondents contend that Alaska's compulsory registration and notification resemble these historical punishments, for they publicize the crime, associate it with his name, and, with the most serious offenders, do so for life.

Any initial resemblance to early punishments is, however, misleading. Punishments such as whipping, pillory, and branding inflicted physical pain and staged a direct confrontation between the offender and the public. Even punishments that lacked the corporal component, such as public shaming, humiliation, and banishment, involved more than the dissemination of information. They either held the person up before his fellow citizens for face-to-face shaming or expelled him from the community. See Earle, supra, at 20, 35-36, 51-52; Massaro, supra, at 1912-1924; Semmes, supra, at 39-40; Blomberg & Lucken, supra, at 30-31. By contrast, the stigma of Alaska's Megan's Law results not from public display for ridicule and shaming but from the dissemination of accurate information about a criminal record, most of which is already public. Our system does not treat dissemination of truthful information in furtherance of a legitimate governmental objective as punishment. On the contrary, our criminal law tradition insists on public indictment, public trial, and public imposition of sentence. Transparency is essential to maintaining public respect for the criminal justice system, ensuring its integrity, and protecting the rights of the accused. The publicity may cause adverse consequences for the convicted defendant, running from mild personal embarrassment to social ostracism. In contrast to the colonial shaming punishments, however, the State does not make the publicity and the resulting stigma an integral part of the objective of the regulatory scheme.

The fact that Alaska posts the information on the Internet does not alter our conclusion. It must be acknowledged that notice of a criminal conviction subjects the offender to public shame, the humiliation increasing in proportion to the extent of the publicity. And the geographic reach of the Internet is greater than anything which could have been designed in colonial times. These facts do not render Internet notification punitive. The purpose and the principal effect of notification are to inform the public for its own safety, not to humiliate the offender. Widespread public access is necessary for the efficacy of the scheme, and the attendant humiliation is but a collateral consequence of a valid regulation.

The State's Web site does not provide the public with means to shame the offender by, say, posting comments underneath his record. An individual seeking the information must take the initial step of going to the Department of Public Safety's Web site, proceed to the sex offender registry, and then look up the desired information. The process is more analogous to a visit to an official archive of criminal records than it is to a scheme forcing an offender to appear in public with some visible badge of past criminality. The Internet makes the document search more efficient, cost effective, and convenient for Alaska's citizenry.

We next consider whether the Act subjects respondents to an "affirmative disability or restraint." Mendoza-Martinez, supra, at 168. Here, we inquire how the effects of the Act are felt by those subject to it. If the disability or restraint is minor and indirect, its effects are unlikely to be punitive.

The Act imposes no physical restraint, and so does not resemble the punishment of imprisonment, which is the paradigmatic affirmative disability or restraint. Hudson, 522 U. S., at 104. The Act's obligations are less harsh than the sanctions of occupational debarment, which we have held to be nonpunitive. See ibid. (forbidding further participation in the banking industry); De Veau v. Braisted, 363 U.S. 144 (1960) (forbidding work as a union official); Hawker v. New York, 170 U.S. 189 (1898) (revocation of a medical license). The Act does not restrain activities sex offenders may pursue but leaves them free to change jobs or residences.

The Court of Appeals sought to distinguish Hawker and cases which have followed it on the grounds that the disability at issue there was specific and "narrow," confined to particular professions, whereas "the procedures employed under the Alaska statute are likely to make [respondents] completely unemployable" because "employers will not want to risk loss of business when the public learns that they have hired sex offenders." 259 F. 3d, at 988. This is conjecture. Landlords and employers could conduct background checks on the criminal records of prospective employees or tenants even with the Act not in force. The record in this case contains no evidence that the Act has led to substantial occupational or housing disadvantages for former sex offenders that would not have otherwise occurred through the use of routine background checks by employers and landlords. The Court of Appeals identified only one incident from the 7-year history of Alaska's law where a sex offender suffered community hostility and damage to his business after the information he submitted to the registry became public. Id., at 987-988. This could have occurred in any event, because the information about the individual's conviction was already in the public domain.

Although the public availability of the information may have a lasting and painful impact on the convicted sex offender, these consequences flow not from the Act's registration and dissemination provisions, but from the fact of conviction, already a matter of public record. The State makes the facts underlying the offenses and the resulting convictions accessible so members of the public can take the precautions they deem necessary before dealing with the registrant.

The Court of Appeals reasoned that the requirement of periodic updates imposed an affirmative disability. In reaching this conclusion, the Court of Appeals was under a misapprehension, albeit one created by the State itself during the argument below, that the offender had to update the registry in person. Id., at 984, n. 4. The State's representation was erroneous. The Alaska statute, on its face, does not require these updates to be made in person. And, as respondents conceded at the oral argument before us, the record contains no indication that an in-person appearance requirement has been imposed on any sex offender subject to the Act. Tr. of Oral Arg. 26-28.

The Court of Appeals held that the registration system is parallel to probation or supervised release in terms of the restraint imposed. 259 F. 3d, at 987. This argument has some force, but, after due consideration, we reject it. Probation and supervised release entail a series of mandatory conditions and allow the supervising officer to seek the revocation of probation or release in case of infraction. See generally Johnson v. United States, 529 U.S. 694 (2000); Griffin v. Wisconsin, 483 U.S. 868 (1987). By contrast, offenders subject to the Alaska statute are free to move where they wish and to live and work as other citizens, with no supervision. Although registrants must inform the authorities after they change their facial features (such as growing a beard), borrow a car, or seek psychiatric treatment, they are not required to seek permission to do so. A sex offender who fails to comply with the reporting requirement may be subjected to a criminal prosecution for that failure, but any prosecution is a proceeding separate from the individual's original offense. Whether other constitutional objections can be raised to a mandatory reporting requirement, and how those questions might be resolved, are concerns beyond the scope of this opinion. It suffices to say the registration requirements make a valid regulatory program effective and do not impose punitive restraints in violation of the Ex Post Facto Clause.

The State concedes that the statute might deter future crimes. Respondents seize on this proposition to argue that the law is punitive, because deterrence is one purpose of punishment. Brief for Respondents 37. This proves too much. Any number of governmental programs might deter crime without imposing punishment. "To hold that the mere presence of a deterrent purpose renders such sanctions `criminal' ... would severely undermine the Government's ability to engage in effective regulation." Hudson, supra, at 105; see also Ursery, 518 U. S., at 292; 89 Firearms, 465 U. S., at 364.

The Court of Appeals was incorrect to conclude that the Act's registration obligations were retributive because "the length of the reporting requirement appears to be measured by the extent of the wrongdoing, not by the extent of the risk posed." 259 F. 3d, at 990. The Act, it is true, differentiates between individuals convicted of aggravated or multiple offenses and those convicted of a single nonaggravated offense. Alaska Stat. § 12.63.020(a)(1) (2000). The broad categories, however, and the corresponding length of the reporting requirement, are reasonably related to the danger of recidivism, and this is consistent with the regulatory objective.

The Act's rational connection to a nonpunitive purpose is a "[m]ost significant" factor in our determination that the statute's effects are not punitive. Ursery, supra, at 290. As the Court of Appeals acknowledged, the Act has a legitimate nonpunitive purpose of "public safety, which is advanced by alerting the public to the risk of sex offenders in their communit[y]." 259 F. 3d, at 991. Respondents concede, in turn, that "this alternative purpose is valid, and rational." Brief for Respondents 38. They contend, however, that the Act lacks the necessary regulatory connection because it is not "narrowly drawn to accomplish the stated purpose." Ibid. A statute is not deemed punitive simply because it lacks a close or perfect fit with the nonpunitive aims it seeks to advance. The imprecision respondents rely upon does not suggest that the Act's nonpunitive purpose is a "sham or mere pretext." Hendricks, 521 U. S., at 371 (KENNEDY, J., concurring).

In concluding the Act was excessive in relation to its regulatory purpose, the Court of Appeals relied in large part on two propositions: first, that the statute applies to all convicted sex offenders without regard to their future dangerousness; and, second, that it places no limits on the number of persons who have access to the information. 259 F. 3d, at 991-992. Neither argument is persuasive.

Alaska could conclude that a conviction for a sex offense provides evidence of substantial risk of recidivism. The legislature's findings are consistent with grave concerns over the high rate of recidivism among convicted sex offenders and their dangerousness as a class. The risk of recidivism posed by sex offenders is "frightening and high." McKune v. Lile, 536 U.S. 24, 34 (2002); see also id., at 33 ("When convicted sex offenders reenter society, they are much more likely than any other type of offender to be rearrested for a new rape or sexual assault" (citing U. S. Dept. of Justice, Bureau of Justice Statistics, Sex Offenses and Offenders 27 (1997); U. S. Dept. of Justice, Bureau of Justice Statistics, Recidivism of Prisoners Released in 1983, p. 6 (1997))).

The Ex Post Facto Clause does not preclude a State from making reasonable categorical judgments that conviction of specified crimes should entail particular regulatory consequences. We have upheld against ex post facto challenges laws imposing regulatory burdens on individuals convicted of crimes without any corresponding risk assessment. See De Veau, 363 U. S., at 160; Hawker, 170 U. S., at 197. As stated in Hawker: "Doubtless, one who has violated the criminal law may thereafter reform and become in fact possessed of a good moral character. But the legislature has power in cases of this kind to make a rule of universal application ...." Ibid. The State's determination to legislate with respect to convicted sex offenders as a class, rather than require individual determination of their dangerousness, does not make the statute a punishment under the Ex Post Facto Clause.

Our decision in Hendricks, on which respondents rely, Brief for Respondents 39, is not to the contrary. The State's objective in Hendricks was involuntary (and potentially indefinite) confinement of "particularly dangerous individuals." 521 U. S., at 357-358, 364. The magnitude of the restraint made individual assessment appropriate. The Act, by contrast, imposes the more minor condition of registration. In the context of the regulatory scheme the State can dispense with individual predictions of future dangerousness and allow the public to assess the risk on the basis of accurate, nonprivate information about the registrants' convictions without violating the prohibitions of the Ex Post Facto Clause.

The duration of the reporting requirements is not excessive. Empirical research on child molesters, for instance, has shown that, "[c]ontrary to conventional wisdom, most reoffenses do not occur within the first several years after release," but may occur "as late as 20 years following release." National Institute of Justice, R. Prentky, R. Knight, & A. Lee, U. S. Dept. of Justice, Child Sexual Molestation: Research Issues 14 (1997).

The Court of Appeals' reliance on the wide dissemination of the information is also unavailing. The Ninth Circuit highlighted that the information was available "world-wide" and "[b]roadcas[t]" in an indiscriminate manner. 259 F. 3d, at 992. As we have explained, however, the notification system is a passive one: An individual must seek access to the information. The Web site warns that the use of displayed information "to commit a criminal act against another person is subject to criminal prosecution." http://www.dps.state.ak.us/nSorcr/asp/ (as visited Jan. 17, 2003) (available in the Clerk of Court's case file). Given the general mobility of our population, for Alaska to make its registry system available and easily accessible throughout the State was not so excessive a regulatory requirement as to become a punishment. See D. Schram & C. Milloy, Community Notification: A Study of Offender Characteristics and Recidivism 13 (1995) (38% of recidivist sex offenses in the State of Washington took place in jurisdictions other than where the previous offense was committed).

The excessiveness inquiry of our ex post facto jurisprudence is not an exercise in determining whether the legislature has made the best choice possible to address the problem it seeks to remedy. The question is whether the regulatory means chosen are reasonable in light of the nonpunitive objective. The Act meets this standard.

The two remaining Mendoza-Martinez factors — whether the regulation comes into play only on a finding of scienter and whether the behavior to which it applies is already a crime — are of little weight in this case. The regulatory scheme applies only to past conduct, which was, and is, a crime. This is a necessary beginning point, for recidivism is the statutory concern. The obligations the statute imposes are the responsibility of registration, a duty not predicated upon some present or repeated violation.

Our examination of the Act's effects leads to the determination that respondents cannot show, much less by the clearest proof, that the effects of the law negate Alaska's intention to establish a civil regulatory scheme. The Act is nonpunitive, and its retroactive application does not violate the Ex Post Facto Clause. The judgment of the Court of Appeals for the Ninth Circuit is reversed, and the case is remanded for further proceedings consistent with this opinion.

It is so ordered.

JUSTICE THOMAS, concurring.

I join the Court's opinion upholding the Alaska Sex Offender Registration Act (ASORA) against ex post facto challenge. I write separately, however, to reiterate that "there is no place for [an implementation-based] challenge" in our ex post facto jurisprudence. Seling v. Young, 531 U.S. 250, 273 (2001) (THOMAS, J., concurring in judgment). Instead, the determination whether a scheme is criminal or civil must be limited to the analysis of the obligations actually created by statute. See id., at 273-274 ("[T]o the extent that the conditions result from the fact that the statute is not being applied according to its terms, the conditions are not the effect of the statute, but rather the effect of its improper implementation"). As we have stated, the categorization of a proceeding as civil or criminal is accomplished by examining "the statute on its face." Hudson v. United States, 522 U.S. 93, 100 (1997) (internal quotation marks omitted).

In this case, ASORA does not specify a means of making registry information available to the public. It states only that

By considering whether Internet dissemination renders ASORA punitive, the Court has strayed from the statute. With this qualification, I concur.

JUSTICE SOUTER, concurring in the judgment.

I agree with the Court that Alaska's Sex Offender Registration Act does not amount to an ex post facto law. But the majority comes to that conclusion by a different path from mine, and I concur only in the judgment.

As the Court says, our cases have adopted a two-step enquiry to see whether a law is punitive for purposes of various constitutional provisions including the Ex Post Facto Clause. At the first step in applying the so-called Kennedy-Ward test, we ask whether the legislature intended a civil or criminal consequence; at the second, we look behind the legislature's preferred classification to the law's substance, focusing on its purpose and effects. See United States v. Ward, 448 U.S. 242, 248-249 (1980); Kennedy v. Mendoza-Martinez, 372 U.S. 144, 168-169 (1963). We have said that "`only the clearest proof'" that a law is punitive based on substantial factors will be able to overcome the legislative categorization. Ward, supra, at 249 (quoting Flemming v. Nestor, 363 U.S. 603, 617 (1960)). I continue to think, however, that this heightened burden makes sense only when the evidence of legislative intent clearly points in the civil direction. See Hudson v. United States, 522 U.S. 93, 113-114 (1997) (SOUTER, J., concurring in judgment). This means that for me this is a close case, for I not only agree with the Court that there is evidence pointing to an intended civil characterization of the Act, but also see considerable evidence pointing the other way.

The Act does not expressly designate the requirements imposed as "civil," a fact that itself makes this different from our past cases, which have relied heavily on the legislature's stated label in finding a civil intent. See Hudson, supra, at 103; Kansas v. Hendricks, 521 U.S. 346, 361 (1997); Allen v. Illinois, 478 U.S. 364, 368 (1986). The placement of the Act in the State's code, another important indicator, see Hendricks, supra, at 361, also leaves matters in the air, for although the section establishing the registry is among the code's health and safety provisions, which are civil, see Alaska Stat. § 18.65.087 (2000), the section requiring registration occurs in the title governing criminal procedure, see § 12.63.010. What is more, the legislature made written notification of the requirement a necessary condition of any guilty plea, see Alaska Rule Crim. Proc. 11(c)(4) (2002), and, perhaps most significant, it mandated a statement of the requirement as an element of the actual judgment of conviction for covered sex offenses, see Alaska Stat. § 12.55.148 (2000); Alaska Rule Crim. Proc. 32(c) (2002). Finally, looking to enforcement, see Hudson, supra, at 103, offenders are obliged, at least initially, to register with state and local police, see §§ 12.63.010(b), (c), although the actual information so obtained is kept by the State's Department of Public Safety, a regulatory agency, see § 18.65.087(a). These formal facts do not force a criminal characterization, but they stand in the way of asserting that the statute's intended character is clearly civil.

The substantial indicators relevant at step two of the Kennedy-Ward analysis likewise point in different directions. To start with purpose, the Act's legislative history shows it was designed to prevent repeat sex offenses and to aid the investigation of reported offenses. See 1994 Alaska Sess. Laws ch. 41, § 1; Brief for Petitioners 26, n. 13. Ensuring public safety is, of course, a fundamental regulatory goal, see, e. g., United States v. Salerno, 481 U.S. 739, 747 (1987), and this objective should be given serious weight in the analyses. But, at the same time, it would be naive to look no further, given pervasive attitudes toward sex offenders, see infra this page and 110, n. See Weaver v. Graham, 450 U.S. 24, 29 (1981) (Ex Post Facto Clause was meant to prevent "arbitrary and potentially vindictive legislation"). The fact that the Act uses past crime as the touchstone, probably sweeping in a significant number of people who pose no real threat to the community, serves to feed suspicion that something more than regulation of safety is going on; when a legislature uses prior convictions to impose burdens that outpace the law's stated civil aims, there is room for serious argument that the ulterior purpose is to revisit past crimes, not prevent future ones. See Kennedy, supra, at 169.

That argument can claim support, too, from the severity of the burdens imposed. Widespread dissemination of offenders' names, photographs, addresses, and criminal history serves not only to inform the public but also to humiliate and ostracize the convicts. It thus bears some resemblance to shaming punishments that were used earlier in our history to disable offenders from living normally in the community. See, e. g., Massaro, Shame, Culture, and American Criminal Law, 89 Mich. L. Rev. 1880, 1913 (1991). While the Court accepts the State's explanation that the Act simply makes public information available in a new way, ante, at 99, the scheme does much more. Its point, after all, is to send a message that probably would not otherwise be heard, by selecting some conviction information out of its corpus of penal records and broadcasting it with a warning. Selection makes a statement, one that affects common reputation and sometimes carries harsher consequences, such as exclusion from jobs or housing, harassment, and physical harm.*

To me, the indications of punitive character stated above and the civil indications weighed heavily by the Court are in rough equipoise. Certainly the formal evidence of legislative intent does not justify requiring the "`clearest proof'" of penal substance in this case, see Hudson, 522 U. S., at 113-114 (SOUTER, J., concurring in judgment), and the substantial evidence does not affirmatively show with any clarity that the Act is valid. What tips the scale for me is the presumption of constitutionality normally accorded a State's law. That presumption gives the State the benefit of the doubt in close cases like this one, and on that basis alone I concur in the Court's judgment.

JUSTICE STEVENS, dissenting in No. 01-729 and concurring in the judgment in No. 01-1231.*

These two cases raise questions about statutes that impose affirmative obligations on convicted sex offenders. The question in No. 01-729 is whether the Alaska Sex Offender Registration Act is an ex post facto law, and in No. 01-1231 it is whether Connecticut's similar law violates the Due Process Clause.

The Court's opinions in both cases fail to decide whether the statutes deprive the registrants of a constitutionally protected interest in liberty. If no liberty interest were implicated, it seems clear that neither statute would raise a colorable constitutional claim. Cf. Meachum v. Fano, 427 U.S. 215 (1976). Proper analysis of both cases should therefore begin with a consideration of the impact of the statutes on the registrants' freedom.

The statutes impose significant affirmative obligations and a severe stigma on every person to whom they apply. In Alaska, an offender who has served his sentence for a single, nonaggravated crime must provide local law enforcement authorities with extensive personal information — including his address, his place of employment, the address of his employer, the license plate number and make and model of any car to which he has access, a current photo, identifying features, and medical treatment — at least once a year for 15 years. If one has been convicted of an aggravated offense or more than one offense, he must report this same information at least quarterly for life. Moreover, if he moves, he has one working day to provide updated information. Registrants may not shave their beards, color their hair, change their employer, or borrow a car without reporting those events to the authorities. Much of this registration information is placed on the Internet. In Alaska, the registrant's face appears on a webpage under the label "Registered Sex Offender." His physical description, street address, employer address, and conviction information are also displayed on this page.

The registration and reporting duties imposed on convicted sex offenders are comparable to the duties imposed on other convicted criminals during periods of supervised release or parole. And there can be no doubt that the "[w]idespread public access," ante, at 99 (opinion in No. 01-729), to this personal and constantly updated information has a severe stigmatizing effect. See Brief for the Office of the Public Defender for the State of New Jersey et al. as Amici Curiae 7-21 (providing examples of threats, assaults, loss of housing, and loss of jobs experienced by sex offenders after their registration information was made widely available). In my judgment, these statutes unquestionably affect a constitutionally protected interest in liberty. Cf. Wisconsin v. Constantineau, 400 U.S. 433 (1971).

It is also clear beyond peradventure that these unique consequences of conviction of a sex offense are punitive. They share three characteristics, which in the aggregate are not present in any civil sanction. The sanctions (1) constitute a severe deprivation of the offender's liberty, (2) are imposed on everyone who is convicted of a relevant criminal offense, and (3) are imposed only on those criminals. Unlike any of the cases that the Court has cited, a criminal conviction under these statutes provides both a sufficient and a necessary condition for the sanction.

To be sure, there are cases in which we have held that it was not punishment and thus not a violation of the Ex Post Facto Clause to deny future privileges to individuals who were convicted of crimes. See, e. g., De Veau v. Braisted, 363 U.S. 144 (1960) (upholding prohibition of convicted felons from working for waterfront unions); Hawker v. New York, 170 U.S. 189 (1898) (upholding prohibition of doctors who had been convicted of a felony from practicing medicine). Those cases are distinguishable because in each the prior conviction was a sufficient condition for the imposition of the burden, but it was not a necessary one. That is, one may be barred from participation in a union because he has not paid fines imposed on him. See NLRB v. Allis-Chalmers Mfg. Co., 388 U.S. 175, 191-192 (1967). And a doctor may not be permitted to practice medicine because she is no longer competent to do so. See, e.g., N. J. Stat. Ann. § 45:1-21 (West Supp. 2002).

Likewise, in Kansas v. Hendricks, 521 U.S. 346 (1997), the Court held that a law that permitted the civil commitment of persons who had committed or had been charged with a sexually violent offense was not an ex post facto law. But the fact that someone had been convicted was not sufficient to authorize civil commitment under Kansas law because Kansas required another proceeding to determine if such a person suffered from a "`mental abnormality or personality disorder which makes the person likely to engage in the predatory acts of sexual violence.'" Id., at 352. Nor was the conviction even a necessary predicate for the commitment. See ibid. (Kansas' civil commitment procedures also applied to individuals charged with a sexually violent offense but found incompetent to stand for trial, or found not guilty by reason of insanity or by reason of mental disease or defect). While one might disagree in other respects with Hendricks, it is clear that a conviction standing alone did not make anyone eligible for the burden imposed by that statute.

No matter how often the Court may repeat and manipulate multifactor tests that have been applied in wholly dissimilar cases involving only one or two of these three aspects of these statutory sanctions, it will never persuade me that the registration and reporting obligations that are imposed on convicted sex offenders and on no one else as a result of their convictions are not part of their punishment. In my opinion, a sanction that (1) is imposed on everyone who commits a criminal offense, (2) is not imposed on anyone else, and (3) severely impairs a person's liberty is punishment.

It is therefore clear to me that the Constitution prohibits the addition of these sanctions to the punishment of persons who were tried and convicted before the legislation was enacted. As the Court recognizes, "recidivism is the statutory concern" that provides the supposed justification for the imposition of such retroactive punishment. Ante, at 105 (opinion in No. 01-729). That is the principal rationale that underlies the "three strikes" statute that the Court has upheld in Ewing v. California, ante, p. 11. Reliance on that rationale here highlights the conclusion that the retroactive application of these statutes constitutes a flagrant violation of the protections afforded by the Double Jeopardy and Ex Post Facto Clauses of the Constitution.

I think it equally clear, however, that the State may impose registration duties and may publish registration information as a part of its punishment of this category of defendants. Looking to the future, these aspects of their punishment are adequately justified by two of the traditional aims of punishment — retribution and deterrence. Moreover, as a matter of procedural fairness, Alaska requires its judges to include notice of the registration requirements in judgments imposing sentences on convicted sex offenders and in the colloquy preceding the acceptance of a plea of guilty to such an offense. See Alaska Rules Crim. Proc. 11(c)(4) and 32(c) (2002). Thus, I agree with the Court that these statutes are constitutional as applied to postenactment offenses.

Accordingly, I would hold that the Alaska statute violates the constitutional prohibition on ex post facto laws. Because I believe registration and publication are a permissible component of the punishment for this category of crimes, however, for those convicted of offenses committed after the effective date of such legislation, there would be no separate procedural due process violation so long as a defendant is provided a constitutionally adequate trial. I therefore concur in the Court's disposition of the Connecticut case, No. 01-1231, and I respectfully dissent from its disposition of the Alaska case, No. 01-729.

JUSTICE GINSBURG, with whom JUSTICE BREYER joins, dissenting.

As JUSTICE SOUTER carefully explains, it is unclear whether the Alaska Legislature conceived of the State's Sex Offender Registration Act as a regulatory measure or as a penal law. See ante, at 107-109 (opinion concurring in judgment). Accordingly, in resolving whether the Act ranks as penal for ex post facto purposes, I would not demand "the clearest proof" that the statute is in effect criminal rather than civil. Instead, guided by Kennedy v. Mendoza-Martinez, 372 U.S. 144 (1963), I would neutrally evaluate the Act's purpose and effects. See id., at 168-169 (listing seven factors courts should consider "[a]bsent conclusive evidence of [legislative] intent as to the penal nature of a statute"); cf. Hudson v. United States, 522 U.S. 93, 115 (1997) (BREYER, J., concurring in judgment) ("[I]n fact if not in theory, the Court has simply applied factors of the Kennedy variety to the matter at hand.").1

Measured by the Mendoza-Martinez factors, I would hold Alaska's Act punitive in effect. Beyond doubt, the Act involves an "affirmative disability or restraint." 372 U. S., at 168. As JUSTICE STEVENS and JUSTICE SOUTER spell out, Alaska's Act imposes onerous and intrusive obligations on convicted sex offenders; and it exposes registrants, through aggressive public notification of their crimes, to profound humiliation and community-wide ostracism. See ante, at 109, and n. (SOUTER, J., concurring in judgment); ante, at 111-112 (STEVENS, J., dissenting in No. 01-729 and concurring in judgment in No. 01-1231).

Furthermore, the Act's requirements resemble historically common forms of punishment. See Mendoza-Martinez, 372 U. S., at 168. Its registration and reporting provisions are comparable to conditions of supervised release or parole; its public notification regimen, which permits placement of the registrant's face on a webpage under the label "Registered Sex Offender," calls to mind shaming punishments once used to mark an offender as someone to be shunned. See ante, at 111-112 (STEVENS, J., dissenting in No. 01-729 and concurring in judgment in No. 01-1231); ante, at 109 (SOUTER, J., concurring in judgment).

Telling too, as JUSTICE SOUTER observes, past crime alone, not current dangerousness, is the "touchstone" triggering the Act's obligations. Ibid. (opinion concurring in judgment); see ante, at 112-113 (STEVENS, J., dissenting in No. 01-729 and concurring in judgment in No. 01-1231). This touchstone adds to the impression that the Act retributively targets past guilt, i.e., that it "revisit[s] past crimes [more than it] prevent[s] future ones." Ante, at 109 (SOUTER, J., concurring in judgment); see Mendoza-Martinez, 372 U. S., at 168.

Tending the other way, I acknowledge, the Court has ranked some laws civil and nonpunitive although they impose significant disabilities or restraints. See, e. g., Flemming v. Nestor, 363 U.S. 603 (1960) (termination of accrued disability benefits payable to deported resident aliens); Kansas v. Hendricks, 521 U.S. 346 (1997) (civil confinement of mentally ill sex offenders). The Court has also deemed some laws nonpunitive despite "punitive aspects." See United States v. Ursery, 518 U.S. 267, 290 (1996).

What ultimately tips the balance for me is the Act's excessiveness in relation to its nonpunitive purpose. See Mendoza-Martinez, 372 U. S., at 169. As respondents concede, see Brief for Respondents 38, the Act has a legitimate civil purpose: to promote public safety by alerting the public to potentially recidivist sex offenders in the community. See ante, at 102-103 (majority opinion). But its scope notably exceeds this purpose. The Act applies to all convicted sex offenders, without regard to their future dangerousness. And the duration of the reporting requirement is keyed not to any determination of a particular offender's risk of reoffending, but to whether the offense of conviction qualified as aggravated. The reporting requirements themselves are exorbitant: The Act requires aggravated offenders to engage in perpetual quarterly reporting, even if their personal information has not changed. See ante, at 90. And meriting heaviest weight in my judgment, the Act makes no provision whatever for the possibility of rehabilitation: Offenders cannot shorten their registration or notification period, even on the clearest demonstration of rehabilitation or conclusive proof of physical incapacitation.2 However plain it may be that a former sex offender currently poses no threat of recidivism, he will remain subject to long-term monitoring and inescapable humiliation.

John Doe I, for example, pleaded nolo contendere to a charge of sexual abuse of a minor nine years before the Alaska Act was enacted. He successfully completed a treatment program, and gained early release on supervised probation in part because of his compliance with the program's requirements and his apparent low risk of reoffense. Brief for Respondents 1. He subsequently remarried, established a business, and was reunited with his family. Ibid. He was also granted custody of a minor daughter, based on a court's determination that he had been successfully rehabilitated. See Doe I v. Otte, 259 F.3d 979, 983 (CA9 2001). The court's determination rested in part on psychiatric evaluations concluding that Doe had "a very low risk of re-offending" and is "not a pedophile." Ibid. (internal quotation marks omitted). Notwithstanding this strong evidence of rehabilitation, the Alaska Act requires Doe to report personal information to the State four times per year, and permits the State publicly to label him a "Registered Sex Offender" for the rest of his life.

Satisfied that the Act is ambiguous in intent and punitive in effect, I would hold its retroactive application incompatible with the Ex Post Facto Clause, and would therefore affirm the judgment of the Court of Appeals.