Clement, J.

D. J. Hilson, Prosecuting Attorney, and Charles F. Justian, Chief Appellate Attorney, for the people.

State Appellate Defender (by Michael L. Mittlestat, Jessica Zimbelman, and Sofia Nelson ) for defendant.

Miriam J. Aukerman, Michael J. Steinberg, Monica Andrade, and Daniel S. Korobkin for the American Civil Liberties Union Fund of Michigan, amicus curiae.

Michigan Appellate Assigned Counsel System (by Bradley R. Hall ) and Warner Norcross + Judd LLP (by Gaëtan Gerville-Réache, Grand Rapids, Nicole A. Samuel, Ashley G. Chrysler, Grand Rapids, and Adam D. Bruski, Midland) for the Criminal Defense Attorneys of Michigan, amicus curiae.

Dana Nessel, Attorney General, Fadwa A. Hammoud, Solicitor General, and Ann M. Sherman, Deputy Solicitor General, for the Attorney General, amicus curiae.

Keith Kushion, Gratiot County Prosecutor, B. Eric Restuccia, Deputy Solicitor General, and Joseph T. Froehlich and Jessica Mullen, Assistant Attorneys General, for the Gratiot County Prosecutor, amicus curiae.

Kimberly A. Thomas for law professors in support of defendant, amicus curiae.

Paul D. Reingold for Safe & Just Michigan, the Michigan Chapter of the National Association of Social Workers, the Michigan Youth Justice Center, the Northwest Initiative, the Professional Advisory Board to the Coalition for a Useful Registry, and the Michigan Collaborative to End Mass Incarceration, amici curiae.

BEFORE THE ENTIRE BENCH (except Welch, J.)

Clement, J.

We are asked to decide whether the retroactive application of Michigan's Sex Offenders Registration Act (SORA), MCL 28.721 et seq. , as amended by 2011 PA 17 and 18 (the 2011 SORA), violates state and federal constitutional prohibitions on ex post facto laws. See U.S. Const., art. I, § 10; Const. 1963, art. 1, § 10. We hold that it does.

I. THE EVOLUTION OF SORA

The Michigan Legislature enacted SORA in 1994 in response to the Jacob Wetterling Crimes Against Children and Sexually Violent Offender Registration Program, 42 USC 14071 , "to better assist law enforcement officers and the people of this state in preventing and protecting against the commission of future criminal sexual acts by convicted sex offenders," MCL 28.721a. This first version of SORA created a confidential database accessible only to law enforcement; it required persons convicted of certain sex offenses to register and notify law enforcement of address changes. MCL 28.725(1), as enacted by 1994 PA 295. Since then, the Legislature has amended the act several times, altering both the nature of the registry and the requirements imposed by it. Defendant alleges that these changes transformed SORA from a regulatory scheme, as it existed in 1996, into a punishment scheme by the time of his failure-to-register conviction in 2012, such that the retroactive application of those provisions to him violated the Ex Post Facto Clauses of the Michigan and United States Constitutions.

The registry became accessible to the public in 1997, when the Legislature required law enforcement to make the registry available for in-person public inspection during business hours. MCL 28.730(2), as amended by 1996 PA 494. Shortly thereafter, in 1999, the Legislature required computerization of the registry and granted law enforcement the authority to make the computerized database available to the public online. MCL 28.728(2), as amended by 1999 PA 85. And in 2006, the Legislature allowed for the registry to send e-mail alerts to any subscribing member of the public when an offender registers within or when a registrant moves into a specified zip code.

As the registry became more accessible to the public, the information registrants were required to provide to law enforcement also expanded. In 2002, the Legislature required registrants to report whenever they enrolled, disenrolled, worked, or volunteered at an institution of higher education. MCL 28.724a, as amended by 2002 PA 542. Two years later, in 2004, the Legislature directed registrants to provide an updated photograph for addition to the online database. MCL 28.728(3)(c), as amended by 2004 PA 238, effective May 1, 2005. And in 2011, the Legislature required registrants to report more personal information, including employment status, "electronic mail addresses and instant message addresses," vehicle information, and travel schedules. MCL 28.727, as amended by 2011 PA 18. Registrants were required to update law enforcement of these changes within three business days, a substantial shortening of the time frame from the initial 10-day reporting window. MCL 28.725(1), as amended by 2011 PA 17. The updates were also required to be made in person rather than by mail, telephone, or e-mail. Id . The 2011 amendments further added a periodic reporting requirement that instructed registrants to present themselves to law enforcement, in person, one or more times a year, even if registrants had no changes to report. MCL 28.725a(3), as amended by 2011 PA 17.

In addition to the expansion of personal information contained in the database, the Legislature also increased other restrictions and obligations imposed by SORA. Specifically, amendments effective in 2006 created "exclusion zones" that prohibited most registrants from living, working, or "loitering" within 1,000 feet of a school. MCL 28.733 to MCL 28.736, as amended by 2005 PA 121. The Legislature also added an annual registration fee of $50. See MCL 28.725a(6), as amended by 2013 PA 149.

The Legislature also enacted significant structural amendments of SORA in 2011. These amendments categorized registrants into three tiers on the basis of their offenses and based the length of registration on that tier designation. MCL 28.722(k) and MCL 28.722(s) through (u), as amended by 2011 PA 17. With this reclassification came lengthened registration periods, including a lifetime registration requirement for Tier III offenders. MCL 28.725(12), as amended by 2011 PA 17. Registrants’ tier classifications were also made available on the public database. MCL 28.728(2)(l ), as amended by 2011 PA 18.

Not all amendments burdened registrants; some were ameliorative. Registration requirements were removed for individuals who were under 14 years old at the time of their offense, MCL 28.722(b), as amended by 2011 PA 17, and for individuals who engaged in consensual but unlawful sexual conduct with a minor under certain conditions, MCL 28.722(t)(v ), as amended by 2011 PA 17. Students enrolled in remote-learning programs for higher education were relieved from reporting their education status, MCL 28.724a(6), as amended by 2011 PA 17. And Tier I offenders’ registration information was removed from public access. MCL 28.728(4)(c), as amended by 2011 PA 18.

SORA initially conceived a confidential law enforcement tool to manage registrants’ names and addresses, but by 2012, that tool transformed into a publicly accessible database that imposed significant restrictions on the lives of registrants. It is this transformation that defendant alleges has caused the retroactive application of the 2011 SORA to violate constitutional ex post facto protections.

II. FACTS AND PROCEDURAL HISTORY

In December 1993, defendant pleaded guilty to second-degree criminal sexual conduct (CSC-II), MCL 750.520c. The trial court sentenced defendant to 5 to 15 years’ imprisonment. Two years later, SORA took effect. After defendant's successful completion of parole, defendant failed to comply with SORA requirements. Specifically, in 2012, defendant failed to report his change of residence, his e-mail address, and his purchase of a vehicle within 3 days, contrary to MCL 28.725(1)(a), (f), and (g), as amended by 2011 PA 17.

The prosecutor charged defendant with violating SORA's registration requirements, MCL 28.729(1)(a). Defendant moved to dismiss the charge, arguing that the retroactive application of the 2011 SORA requirements violated the constitutional prohibitions on ex post facto laws. The trial court denied this motion. Defendant ultimately entered a no-contest plea, conditional on his ability to challenge on appeal the constitutionality of the retroactive application of the 2011 SORA. The trial court sentenced defendant to 36 months’ probation, with 12 months’ jail time, but suspended imposition of that sentence during the pendency of defendant's appeal.

Defendant sought leave to appeal in the Court of Appeals, and the Court of Appeals denied defendant's application for lack of merit in the grounds presented. Defendant subsequently sought leave to appeal in this Court. After a period of abeyance for the resolution of related cases, this Court heard oral argument on the application in March 2019. Following oral argument on the application, this Court granted defendant's application for leave to appeal and directed further oral argument as to the following issues:

(1) whether the requirements of the Sex Offenders Registration Act (SORA), MCL 28.721 et seq. , taken as a whole, amount to "punishment" for the purposes of the Ex Post Facto Clauses of the Michigan and United States Constitutions, U.S. Const., art. I, § 10; Const. 1963, art. 1, § 10 ; see People v. Earl , 495 Mich. 33[, 845 N.W.2d 721] (2014), see also Does #1-5 v. Snyder , 834 F.3d 696, 703-706 (C.A. 6, 2016), cert. den. sub. nom. Snyder v. John Does #1-5 , [––– U.S. ––––] 138 S. Ct. 55 (2017) ; (2) if SORA, as a whole, constitutes punishment, whether it became punitive only upon the enactment of a certain provision or group of provisions added after the initial version of SORA was enacted; (3) if SORA only became punitive after a particular enactment, whether a resulting ex post facto violation would be remedied by applying the version of SORA in effect before it transformed into a punishment or whether a different remedy applies, see Weaver v. Graham , 450 U.S. 24, 36 n. 22[ 101 S. Ct. 960, 67 L. Ed. 2d 17] (1981) ("the proper relief ... is to remand to permit the state court to apply, if possible, the law in place when his crime occurred."); (4) if one or more discrete provisions of SORA, or groups of provisions, are found to be ex post facto punishments, whether the remaining provisions can be given effect retroactively without applying the ex post facto provisions, see MCL 8.5 ; (5) what consequences would arise if the remaining provisions could not be given retroactive effect; and (6) whether the answers to these questions require the reversal of the defendant's conviction pursuant to MCL 28.729 for failure to register under SORA.^{[ ]}

Following oral argument, the Legislature enacted a series of amendments of SORA, effective March 24, 2021. 2020 PA 295. This Court subsequently issued an order directing the parties to provide supplemental briefing addressing the effect, if any, of the new legislation on the present case. III. PARALLEL FEDERAL LITIGATION

During defendant's appeal in state court, related litigation has progressed through the federal courts. In 2012, five plaintiffs required to register as Tier III offenders sued Michigan's governor and the director of the Michigan State Police, arguing that the 2011 SORA was unconstitutional on several grounds. In a series of opinions, the district court partially ruled in the plaintiffs’ favor, holding that the 2011 SORA's student-safety zone provisions were unconstitutionally vague, that certain in-person reporting provisions were unconstitutionally vague, that certain in-person reporting provisions violated the First Amendment, and that registrants could not be held strictly liable for violating the 2011 SORA's requirements. However, the district court rejected the remainder of the plaintiffs’ claims, including their argument that the retroactive application of the 2011 SORA violated ex post facto protections.

On appeal, the United States Court of Appeals for the Sixth Circuit disagreed, concluding that the retroactive application of the 2011 SORA did violate constitutional ex post facto provisions. Does #1-5 v. Snyder , 834 F.3d 696, 705-706 (C.A. 6, 2016) ( Does I ). It reasoned that the cumulative punitive effects of the 2011 SORA outweighed the nonpunitive intent of the Legislature such that the retroactive application of the 2011 SORA constituted the retroactive application of punishment in violation of the federal Constitution. Id. Because this holding rendered the 2011 SORA inapplicable to the federal plaintiffs, the Sixth Circuit declined to address the remainder of the issues decided by the district court. Id. at 706. The United States Supreme Court denied certiorari. Snyder v. Does # 1-5 , ––– U.S. ––––, 138 S. Ct. 55, 199 L.Ed.2d 18 (2017).

Shortly after the Sixth Circuit's decision in Does I , six other plaintiffs filed a class-action complaint in the federal district court challenging the constitutionality of the 2011 SORA on the same grounds raised by the Does I plaintiffs. These plaintiffs also noted that although the Does I plaintiffs had received a favorable ruling from the Sixth Circuit on their ex post facto challenge, the state of Michigan had continued to enforce the 2011 SORA against all SORA registrants. Ultimately, the district court ruled for the plaintiffs and entered an order permanently enjoining the state of Michigan from enforcing the unconstitutional provisions of the 2011 SORA identified in Does I against any registrant and from enforcing the 2011 SORA retroactively. Doe v. Snyder , 449 F. Supp. 3d 719, 737-738 (E.D. Mich., 2020) ( Does II ). In so doing, the district court rejected the possibility that portions of the 2011 SORA or an earlier version of SORA could be constitutionally applied retroactively. Id. at 731-735. The district court also rejected the defendants’ request to certify these issues to this Court. Id. at 729-731. The district court directed the parties to draft a proposed judgment and ordered that the judgment would be effective 60 days after its entry. Id. at 739. However, that process was hindered by the global outbreak of the severe acute respiratory disease known as COVID-19. In April 2020, the district court entered an order suspending final judgment "for the duration of the current COVID-19 crisis" but preliminarily enjoining the state from "enforcing registration, verification, school zone, and fee violations of SORA that occurred or may occur from February 14, 2020, until the current crisis has ended ...." Doe v. Snyder , ––– F. Supp. 3d ––––, ––––, 2020 WL 1684156 *2 (E.D. Mich., 2020) (Case No. 16-13137). Proceedings resumed the following year, and in June 2021, the district court issued an order resolving several disagreements regarding the content of a final judgment. Doe v. Snyder , ––– F. Supp. 3d ––––, –––– (E.D. Mich., 2021) (Case No. 16-13137). This order also extended the interim injunction to July 12, 2021, and directed the parties to produce a joint proposed judgment by that time. Id. The court subsequently extended the deadline to July 19, 2021. To date, a final judgment has not been entered.

IV. EX POST FACTO

This Court is asked to determine whether the retroactive application of the 2011 SORA violates federal and state constitutional ex post facto protections. Although the Sixth Circuit in Does I determined that the retroactive application of the 2011 SORA violates federal constitutional ex post facto protections, this Court is not bound by that determination, see Johnson v. VanderKooi , 502 Mich. 751, 764 n. 6, 918 N.W.2d 785 (2018), and the Sixth Circuit's opinion did not assess an ex post facto challenge under our state constitutional law. Both the Michigan and United States Constitutions prohibit ex post facto laws. U.S. Const., art. I, § 10 ; Const. 1963, art. 1, § 10. A law is considered ex post facto if it: "(1) punishes an act that was innocent when the act was committed; (2) makes an act a more serious criminal offense; (3) increases the punishment for a [committed] crime; or (4) allows the prosecution to convict on less evidence." People v. Earl , 495 Mich. 33, 37, 845 N.W.2d 721 (2014). The prohibitions on ex post facto laws "assure that legislative Acts give fair warning of their effect and permit individuals to rely on their meaning" as well as prevent the government from imposing arbitrary and vindictive legislation. Weaver v. Graham , 450 U.S. 24, 28-29, 101 S. Ct. 960, 67 L. Ed. 2d 17 (1981). See also The Federalist No. 44 (Madison) (Rossiter ed., 1961), p. 282 (stating that "ex post facto laws ... are contrary to the first principles of the social compact and to every principle of sound legislation"); The Federalist No. 84 (Hamilton) (Rossiter ed., 1961), pp. 511-512 (observing that ex post facto laws have historically been "the favorite and most formidable instruments of tyranny").

At issue here is the third type of ex post facto law, namely, whether the retroactive application of the 2011 SORA unconstitutionally increases the punishment for defendant's CSC-II conviction. To answer this question, this Court must engage in a two-step inquiry. Earl , 495 Mich. at 38, 845 N.W.2d 721. First, this Court must determine "whether the Legislature intended the statute as a criminal punishment or a civil remedy." Id. If the statute imposes a disability for the purpose of reprimanding the wrongdoer, the Legislature likely intended the statute to be a criminal punishment. Id. However, if the statute imposes a disability to further a legitimate public purpose, the Legislature likely intended the statute to be a civil or regulatory remedy. Id.

If the Legislature intended to impose criminal punishment, the retroactive application of such a statute violates the ex post facto prohibitions, and the inquiry ends. Id. However, if the Legislature intended to impose a civil or regulatory remedy, this Court must then consider "whether the statutory scheme is so punitive either in purpose or effect as to negate the State's intention to deem it civil." Id. (quotation marks, citation, and brackets omitted). To aid in that analysis, the United States Supreme Court has provided that the following nonexhaustive factors are relevant to the inquiry:

Whether the sanction involves an affirmative disability or restraint, whether it has historically been regarded as a punishment, whether it comes into play only on a finding of scienter , whether its operation will promote the traditional aims of punishment—retribution and deterrence, whether the behavior to which it applies is already a crime, whether an alternative purpose to which it may rationally be connected is assignable for it, and whether it appears excessive in relation to the alternative purpose assigned .... [ Kennedy v. Mendoza-Martinez , 372 U.S. 144, 168-169, 83 S. Ct. 554, 9 L. Ed. 2d 644 (1963) (citations omitted). See also Earl , 495 Mich. at 43-44, 845 N.W.2d 721 (noting the Mendoza-Martinez factors as the proper avenue of analysis for this issue).]

Further, the Legislature's manifest intent will be rejected only when "a party challenging the statute provides the clearest proof that the statutory scheme is so punitive either in purpose or effect to negate the State's intention to deem it civil." Kansas v. Hendricks , 521 U.S. 346, 361, 117 S. Ct. 2072, 138 L. Ed. 2d 501 (1997) (quotation marks, citation, and brackets omitted; emphasis added).

In 2003, the United States Supreme Court applied this two-step inquiry when it considered whether Alaska's sex-offender registry statute violated state and federal ex post facto protections. Smith v. Doe , 538 U.S. 84, 123 S. Ct. 1140, 155 L. Ed. 2d 164 (2003). The Alaska registry statute required every convicted sex offender in the state to provide law enforcement with their name, aliases, identifying features, address (including anticipated changes of address), place of employment, date of birth, conviction information, driver's license number, information about vehicles to which they had access, and postconviction history of medical treatment. Id. at 90, 123 S. Ct. 1140. The information regarding driver's license numbers, anticipated changes of addresses, and whether the registrant sought and obtained medical treatment was kept confidential; other information was available to the public online. Id. at 90-91, 123 S. Ct. 1140. The amount of time that a person was required to remain registered with this system was based on the registrant's number and type of convictions. Id. at 90, 123 S. Ct. 1140. Two respondents who pleaded nolo contendere to sexual abuse of a minor before the registry scheme was enacted brought suit under 42 USC 1983, seeking a declaration that the registry statute was unlawful as applied to them because it violated ex post facto protections. Id. at 91, 123 S. Ct. 1140. The United States Supreme Court first determined that the Alaska Legislature had intended its registry law to be nonpunitive, reasoning that the statutory text provided a purpose of public safety; that the statute was codified within Alaska's Health, Safety, and Housing Code; and that the authority to promulgate implementing procedures was vested in the Alaska Department of Public Safety. Id. at 93-96, 123 S. Ct. 1140. It next considered whether the effects of the statutory scheme were so punitive in effect as to negate the Alaska Legislature's nonpunitive intent. Id. at 97, 123 S. Ct. 1140. The Court acknowledged that the Mendoza-Martinez factors were neither exhaustive nor individually dispositive, and the Court identified the following factors as particularly relevant to the case at hand: "whether, in its necessary operation, the regulatory scheme: [1] has been regarded in our history and traditions as a punishment; [2] imposes an affirmative disability or restraint; [3] promotes the traditional aims of punishment; [4] has a rational connection to a nonpunitive purpose; or [5] is excessive with respect to this purpose." Id.

With regard to the first factor, whether the regulatory scheme has been regarded in our history and traditions as punishment, the Court reasoned that sex-offender registries are not traditional means of punishment because they were of relatively recent design. Id. The Court rejected the respondents’ argument that the registry resembled colonial-era shaming punishments because those punishments "involved more than the dissemination of information," and the Court further noted that, although some registrants might experience negative effects because of public access to their information, "[o]ur system does not treat dissemination of truthful information in furtherance of a legitimate governmental objective as punishment." Id. at 98-99, 123 S. Ct. 1140. Finally, the Court observed that the registry was "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." Id. at 99, 123 S. Ct. 1140. With regard to the second factor, whether the regulatory scheme imposes an affirmative disability or restraint, the Court concluded that the registry did not do so because it did not impose any physical restraint. Id. at 100, 123 S. Ct. 1140. It specifically observed that the registry allowed registrants to change jobs or residences as they desired and imposed no requirement to appear in person. Id. at 100-101, 123 S. Ct. 1140.

With regard to the third factor, whether the regulatory scheme promotes traditional aims of punishment, the Court considered Alaska's concession that the registry might deter future crimes but ruled that this concession did not render the registry punitive. Id. at 102, 123 S. Ct. 1140 ("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.") (quotation marks and citation omitted). The Court also held that the registry's lack of individualized risk assessment did not render it retributive, reasoning that the registry logically related the length of reporting requirements to the amount and severity of the registrant's convictions, which was consistent with the registry's regulatory objective of public safety. Id. at 102-103, 123 S. Ct. 1140.

With regard to the fourth factor, whether the regulatory scheme has a rational connection to a nonpunitive purpose, the Court held that the registry was rationally connected to the nonpunitive purpose of public safety because it alerted the public to the risk of sex offenders in their community. Id. at 103-104, 123 S. Ct. 1140. And, regarding the final factor, excessiveness, the Court held that the registry's requirements were not excessive in relation to its nonpunitive goal. Id. at 103-106, 123 S. Ct. 1140. Specifically, the Court reasoned:

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, 122 S.Ct. 2017, 153 L.Ed.2d 47 (2002) ; see also id. , at 33, 122 S.Ct. 2017 ("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))). [ Smith , 538 U.S. at 103, 123 S.Ct. 1140.]

Taking into account all the Mendoza-Martinez factors, the Court concluded that any punitive effects of Alaska's sex-offender registry did not overcome the Alaska Legislature's intent to establish a civil regulatory scheme. Id. at 105-106, 123 S. Ct. 1140. Accordingly, the retroactive application of the registry's requirements did not violate federal constitutional ex post facto protections. Id. at 106, 123 S. Ct. 1140.

Although Michigan's SORA as initially enacted was similar to the Alaska sex-offender registry at issue in Smith , subsequent amendments have imposed additional requirements and prohibitions on registrants, warranting a fresh look at how the 2011 SORA fares under the constitutional ex post facto protections. See, e.g., Doe v. State , 189 P.3d 999, 1017 (Alas., 2008) (wherein the Alaska Supreme Court held that because of intervening amendments of its sex-offender registry that increased requirements and restrictions on registrants, the retroactive application of its sex-offender registry laws violated ex post facto protections). A. LEGISLATIVE PURPOSE

This Court must first consider "whether the Legislature intended the statute as a criminal punishment or a civil remedy." Earl , 495 Mich. at 38, 845 N.W.2d 721. When the Legislature amended SORA in 2002, it included the following statement:

The legislature declares that the sex offenders registration act was enacted pursuant to the legislature's exercise of the police power of the state with the intent to better assist law enforcement officers and the people of this state in preventing and protecting against the commission of future criminal sexual acts by convicted sex offenders. The legislature has determined that a person who has been convicted of committing an offense covered by this act poses a potential serious menace and danger to the health, safety, morals, and welfare of the people, and particularly the children, of this state. The registration requirements of this act are intended to provide law enforcement and the people of this state with an appropriate, comprehensive, and effective means to monitor those persons who pose such a potential danger. [ MCL 28.721a.]

This description indicates that the Legislature's intent in enacting SORA was the promotion of public safety, a nonpunitive goal. Further, SORA is codified in Chapter 28 of the Michigan Compiled Laws rather than Chapter 750, the Michigan Penal Code. However, other aspects of SORA suggest a punitive intent. Although MCL 28.721a describes the Legislature's intent as the promotion of public safety, it does so by seeking to deter future crimes, a traditional penological aim. Further, SORA requirements are imposed as a consequence of a criminal conviction, its requirements are enforced by law enforcement, and violations of its requirements are punishable by criminal conviction. Weighing these characteristics against the Legislature's expression of its intent in MCL 28.721a, we conclude that the Legislature likely intended SORA as a civil regulation rather than a criminal punishment.

B. PUNITIVE EFFECTS

Because we conclude that the Legislature likely intended SORA as a civil regulation, we must now determine "whether the statutory scheme is so punitive either in purpose or effect as to negate the State's intention to deem it civil." Earl , 495 Mich. at 38, 845 N.W.2d 721 (quotation marks, citation, and brackets omitted). Again, a challenging party must provide "the clearest proof" of the statutory scheme's punitive character in order "to [successfully] negate the State's intention to deem it civil." Hendricks , 521 U.S. at 361, 117 S.Ct. 2072 (quotation marks, citation, and brackets omitted). In determining whether defendant has satisfied this burden, we do not examine individual provisions of SORA in isolation but instead assess SORA's punitive effect in light of all the act's provisions when viewed as a whole. See Smith , 538 U.S. at 92, 94, 96-97, 99, 104-105, 123 S.Ct. 1140 ; see also Doe v. State , 167 N.H. 382, 402, 111 A.3d 1077 (2015) (holding that the punitive-effect "inquiry cannot be answered by looking at the effect of any single provision in the abstract"; rather, a court "must consider the effect of all the provisions and their cumulative impact upon the defendant's rights") (quotation marks and citations omitted). We assess in turn each of the Mendoza-Martinez factors that the United States Supreme Court identified as relevant in Smith .

1. HISTORY AND TRADITION

This Mendoza-Martinez factor asks this Court to consider whether SORA has "been regarded in our history and traditions as a form of criminal punishment." Earl , 495 Mich. at 45, 845 N.W.2d 721. Sex-offender registries are of relatively recent origin and so have no direct analogies in this nation's history and traditions. See Smith , 538 U.S. at 97, 123 S.Ct. 1140. However, the 2011 SORA does resemble, in some respects, the traditional punishments of banishment, shaming, and parole.

In regard to banishment, the 2011 SORA's student-safety zones excluded registrants from working, living, or loitering within 1,000 feet of school property. See MCL 28.733 to MCL 28.736, as amended by 2005 PA 121. Unlike traditional banishment, these exclusion zones did not explicitly exile a registrant from the community. See United States v. Ju Toy , 198 U.S. 253, 269-270, 25 S. Ct. 644, 49 L. Ed. 1040 (1905). But they might have effectively banished a registrant from living within the community. For example, in urban areas that host several schools within their geographic borders, the 1,000-foot restriction emanating from each school might have eliminated access to affordable housing. See, e.g., Does I , 834 F.3d at 702 (providing a visual representation of exclusion zones in Grand Rapids). Or, in rural areas with fewer schools but concentrated community areas, the 1,000-foot restriction might have eliminated a registrant's access to employment and resources within the town or city center. And available homeless shelters might have also been encompassed by the 1,000-foot residency restriction. Compare with Smith , 538 U.S. at 101, 123 S.Ct. 1140 (noting that the 2003 Alaska sex-offender registry, which the United States Supreme Court held did not violate ex post facto protections, left registrants "free to move where they wish[ed] and to live and work as other citizens").

The 2011 SORA also resembles the punishment of shaming. The breadth of information available to the public—far beyond a registrant's criminal history—as well as the option for subscription-based notification of the movement of registrants into a particular zip code, increased the likelihood of social ostracism based on registration. While the initial version of SORA might have been "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," Smith , 538 U.S. at 99, 123 S.Ct. 1140, its 2011 iteration contained more personal information and required less effort to access that information. The public-facing registry contained not only information regarding a registrant's criminal conviction but also the registrant's home address, place of employment, sex, race, age, height, weight, hair and eye color, discernible features, and tier classification. When SORA's notification provision was used, members of the public were alerted to this information without active effort on their behalf, in sharp contrast with the endeavor of visiting an official archive for information. Further, a registrant's information could precede his entrance into a community, increasing the likelihood of ostracism. MCL 28.721a itself—the Legislature's statement of its intent in enacting SORA—refers to providing the public, not just law enforcement, with the means to monitor persons with sex-offense convictions, encouraging public participation and engagement with the registry and furthering the stigma of registration. See Doe , 167 N.H. at 406, 111 A.3d 1077 ("Placing offenders’ pictures and information online serves to notify the community, but also holds them out for others to shame or shun."). As with banishment, however, the 2011 SORA does not perfectly resemble the traditional punishment of shaming. See Smith , 538 U.S. at 97-98, 123 S.Ct. 1140 (describing traditional shaming punishments such as requiring offenders to stand in public with signs describing their crimes or branding offenders in order to inflict permanent stigma). The 2011 SORA did not provide a conduit for the public to directly criticize and shame registrants—as it would have, for example, if it provided an online forum or area for comments in addition to the online registry.

Finally, the 2011 SORA also resembles parole. Although registrants need not have sought permission to make life changes, they were not free to live and work where they desired because of the student-safety zones. Registrants, like parolees, were required to periodically report in person to law enforcement. See MCL 28.725a(3), as amended by 2011 PA 17; MCL 28.725(1), as amended by 2011 PA 17. They were also required to pay registration fees. See MCL 28.725a(6), as amended by 2011 PA 17. Failure to comply with SORA's requirements, like the failure to comply with parole conditions, potentially subjects the offender to imprisonment. See MCL 28.729(1), as amended by 2011 PA 18. Further, as with parole, a law enforcement officer at any time could have investigated a registrant's status based on an anonymous tip. The 2011 SORA thus imposed a significant amount of supervision by the state on registrants. This amount of supervision differentiates the 2011 SORA from the 2003 Alaska sex-offender registry, which the United States Supreme Court held did not resemble parole because registrants were "free to move where they wish and to live and work as other citizens" and because registrants were not required to make periodic updates to law enforcement in person. Smith , 538 U.S. at 101-102, 123 S.Ct. 1140. Neither of these characteristics is true of the 2011 SORA.

In conclusion, the 2011 SORA bears significant resemblance to the traditional punishments of banishment, shaming, and parole because of its limitations on residency and employment, publication of information and encouragement of social ostracism, and imposition of significant state supervision. 2. AFFIRMATIVE DISABILITY OR RESTRAINT

This Mendoza-Martinez factor asks this Court to "inquire how the effects of" the 2011 SORA "are felt by those subject to it." Smith , 538 U.S. at 99-100, 123 S.Ct. 1140. "If the disability or restraint is minor and indirect, its effects are unlikely to be punitive." Id. at 100, 123 S. Ct. 1140.

Imprisonment is the "paradigmatic" affirmative restraint, id. , and the 2011 SORA ensured adherence to its many requirements on the potential for imposition of imprisonment. Although SORA has always contained such a penalty provision, the conditions that a registrant must satisfy to avoid incarceration have increased. See Doe , 167 N.H. at 403, 111 A.3d 1077 (explaining that courts have found sex-offender registry requirements "to be amplified" when "the failure to comply with the requirements could result in harsh prosecution and penalties, such as a fine or imprisonment"). Even those who adhered faithfully to the 2011 SORA's requirements were subject to onerous burdens. As discussed earlier, the 2011 SORA affirmatively barred registrants from living, working, and loitering in large regions of the state through the student-safety zones. The application of these exclusionary zones also had substantial collateral consequences, including limiting access to public transportation, employment opportunities, educational opportunities, resources like counseling and mental-health treatment, and medical care such as residential nursing homes. The 2011 SORA's definition of "loiter"—"to remain for a period of time and under circumstances that a reasonable person would determine is for the primary purpose of observing or contacting minors," MCL 28.733(b), as amended by 2005 PA 121—was also arguably broad enough to encompass parenting activities such as attending parent-teacher conferences, attending student sporting events, or transporting children to school. Further, the student-safety zones also risked preventing registrants from establishing a permanent home, given that registrants were not excepted from the residency ban if a school opened within 1,000 feet of their established home. See MCL 28.735, as amended by 2005 PA 322.

The in-person reporting requirements imposed by former MCL 28.725(1) also imposed affirmative disabilities on registrants. Upon numerous life events, including moving residences, changing employment, changing educational status, changing names, making plans to reside outside of the home for more than 7 days, and changing vehicles, registrants were required to provide an in-person update to law enforcement within three days. MCL 28.725(1), as amended by 2011 PA 17. Particularly onerous was the requirement in former MCL 28.725(1)(f) of immediate in-person reporting when a registrant established "any electronic mail or instant message address, or any other designations used in internet communications or postings." Given the ubiquity of the Internet in daily life, this requirement might have been triggered dozens of times within a year. In addition to the in-person reporting requirements triggered by life events, each registrant was required to make periodic in-person reports to law enforcement: for Tier I offenders, yearly; for Tier II offenders, semiannually; and for Tier III offenders, quarterly. MCL 28.725a(3), as amended by 2011 PA 17. Cumulatively, these frequent in-person reports imposed a burden on registrants, especially for those who might have had difficulty traveling to make the reports—such as those who did not have access to public transportation, did not have the financial resources necessary for private or public transportation, or had health or accessibility issues that would have impeded transportation. See State v. Letalien , 985 A.2d 4, 24-25, 2009 ME 130 (2009) ("[I]t belies common sense to suggest that a newly imposed lifetime obligation to report to a police station every ninety days to verify one's identification, residence, and school, and to submit to fingerprinting and provide a current photograph, is not a substantial disability or restraint on the free exercise of individual liberty.").

In sum, the 2011 SORA imposed onerous restrictions on registrants by restricting their residency and employment, and it also imposed significant affirmative obligations by requiring extensive in-person reporting.

3. TRADITIONAL AIMS OF PUNISHMENT

This Mendoza-Martinez factor asks the Court to consider whether the 2011 SORA promotes the traditional aims of punishment: retribution and specific and general deterrence. See Earl , 495 Mich. at 46, 845 N.W.2d 721.

As the prosecutor concedes, SORA promotes the aim of deterrence. Deterrence is necessarily encompassed by SORA's stated purpose of "preventing and protecting against the commission of future criminal sexual acts by convicted sex offenders." MCL 28.721a. The extensive requirements of the 2011 SORA also generally deterred potential offenders by increasing the resultant consequences of sexual predation. Yet the aim of deterrence alone does not render a statute punitive. See Smith , 538 U.S. at 102, 123 S.Ct. 1140. As the United States Supreme Court has reasoned, "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." Id. (quotation marks and citation omitted). However, the deterrent effect of the 2011 SORA is not merely an indirect consequence, incidental to its regulatory function, but instead a main feature of the statutory scheme. See MCL 28.721a.

The 2011 SORA also supports the aim of retribution. The 2011 SORA was imposed on offenders for the sole fact of their prior offenses and made no individualized determination of the dangerousness of each registrant, indicating that SORA's restrictions were retribution for past offenses rather than regulations to prevent future offenses. See Smith , 538 U.S. at 109, 123 S.Ct. 1140 (Souter, J., concurring in the judgment) ("The fact that the Act uses past crimes 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.").

In sum, because the 2011 SORA aimed to protect the public through deterrence and because its restrictions appear retributive, the 2011 SORA promotes the traditional aims of punishment.

4. CONNECTION TO NONPUNITIVE PURPOSE

Next, this Court must consider whether the 2011 SORA has a rational connection to a nonpunitive purpose. See Earl , 495 Mich. at 46, 845 N.W.2d 721. A rational connection is all that is required; "[a] statute is not deemed punitive simply because it lacks a close or perfect fit with the nonpunitive aims it seeks to advance." Smith , 538 U.S. at 103, 123 S.Ct. 1140 (opinion of the Court).

Again, the asserted goal of the Legislature is to "provide law enforcement and the people of this state with an appropriate, comprehensive, and effective means to monitor those persons" who have committed a specified sex offense and who are therefore considered to "pose[ ] a potential serious menace and danger to the health, safety, morals, and welfare of the people ... of this state." MCL 28.721a. The protection of citizens from potentially dangerous sex offenders is "a compelling state interest in furtherance of the state's police powers." Letalien , 985 A.2d at 22. The 2011 SORA, by identifying potentially recidivist sex offenders and alerting the public, seeks to further the nonpunitive purpose of public safety. Accordingly, given the low bar of rationality, the 2011 SORA is connected to a nonpunitive purpose. 5. EXCESSIVENESS

The final Mendoza-Martinez factor to be assessed is "whether the regulatory means chosen are reasonable in light of the nonpunitive objective." Smith , 538 U.S. at 105, 123 S.Ct. 1140 (opinion of the Court). Similar to the rational-connection determination, the lynchpin of this analysis is "reasonableness," not "whether the legislature has made the best choice possible to address the problem it seeks to remedy." Id.

The Legislature's asserted nonpunitive goal was based on the Legislature's determination that "a person who has been convicted of committing an offense covered by [SORA] poses a potential serious menace and danger to the health, safety, morals, and welfare of the people, and particularly the children, of this state." MCL 28.721a. Central to our inquiry, then, is whether the 2011 SORA is a reasonable means of protecting the public from sex offenders who allegedly pose such a "potential serious menace." Id. Defendant—as well as his similarly situated counterparts throughout the nation—endeavors to demonstrate that the dangerousness of sex offenders has been historically overblown and that, in fact, sex offenders are actually less likely to recidivate than other offenders. Further, he argues that sex-offender registries have dubious efficacy in achieving their professed goals of decreasing recidivism. A growing body of research supports these propositions. See, e.g., United States Department of Justice, Alper & Durose, Recidivism of Sex Offenders Released from State Prison: A 9-Year Follow-Up (2005-2014) (May 2019), available at < https://www.bjs.gov/content/pub/pdf/rsorsp9yfu0514.pdf> (accessed June 4, 2021) [https://perma.cc/U9HY-MJ7F] (concluding that sex offenders are less likely than other offenders to be rearrested for any crime); Huebner et al., An Evaluation of Sex Offender Residency Restrictions in Michigan and Missouri (July 1, 2013), p. 72, available at < https://www.ncjrs.gov/pdffiles1/nij/grants/242952.pdf> (accessed June 4, 2021) [https://perma.cc/D9K4-CV5P] (concluding that residency restrictions "are unlikely to mitigate or reduce the risk of recidivism among sex offenders"); Prescott & Rockoff, Do Sex Offender Registration and Notification Laws Affect Criminal Behavior? , 54 J. L. & Econ. 161, 192 (2011) (concluding that notification requirements in a typical sex-offender registry "effectively increases the number of sex offenses by more than 1.57 percent," likely "because of the social and financial costs associated with the public release of their criminal history and personal information"). For our limited purpose in examining the potential excessiveness of the 2011 SORA in regard to its public-safety purpose, these studies demonstrate that, at minimum, the 2011 SORA's efficacy is unclear.

Given the uncertainty of the 2011 SORA's efficacy, the restraints it imposed were excessive. Over 40,000 registrants were subject to the 2011 SORA's requirements without any individualized assessment of their risk of recidivism. The duration of an offender's reporting requirement was based solely on the offender's conviction and not the danger he individually posed to the community. Registrants remained subject to SORA—including the stigma of having been branded a potentially violent menace by the state—long after they had completed their sentence, probation, and any required treatment. All registrants were excluded from residing, working, and loitering within 1,000 feet of a school, even those whose offenses did not involve children and even though most sex offenses involving children are perpetrated by a person already known to the child. As described, this restriction placed significant burdens on registrants’ ability to find affordable housing, obtain employment, and participate as a member of the community. Registrants were also required to make frequent in-person reports to law enforcement upon minor life changes and regular in-person reports—sometimes multiple times a year—even when no information had changed. These demanding and intrusive requirements, imposed uniformly on all registrants regardless of an individual's risk of recidivism, were excessive in comparison to SORA's asserted public-safety purpose.

Considering the Mendoza-Martinez factors cumulatively, the 2011 SORA's aggregate punitive effects negate the state's intention to deem it a civil regulation. See Earl , 495 Mich. at 38, 845 N.W.2d 721. Accordingly, the retroactive imposition of the 2011 SORA increases registrants’ punishment for their committed offenses in violation of federal and state constitutional prohibitions on ex post facto laws.

V. REMEDY

Having concluded that the retroactive application of the 2011 SORA violates constitutional ex post facto provisions, we turn to the issue of remedy. Although the 2011 SORA did not contain a general severability provision, Michigan has a legislative preference for severability, as expressed in MCL 8.5 :

In the construction of the statutes of this state the following rules shall be observed, unless such construction would be inconsistent with the manifest intent of the Legislature, that is to say:

If any portion of an act or the application thereof to any person or circumstances shall be found to be invalid by a court, such invalidity shall not affect the remaining portions or applications of the act which can be given effect without the invalid portion or application, provided such remaining portions are not determined by the court to be inoperable, and to this end acts are declared to be severable. [See also Blank v. Dep't of Corrections , 462 Mich. 103, 122-123, 611 N.W.2d 530 (2000) (stating that the "general rule" regarding laws determined to be unconstitutional "favors severability").]

Because this Court has found that the retroactive application of the 2011 SORA is unconstitutional, this Court must now consider whether "the remaining portions or applications of the act ... can be given effect without the invalid ... application ...." MCL 8.5. This Court's conclusion does not affect the prospective application of the 2011 SORA, and so we must consider only whether certain provisions of the 2011 SORA can be given retroactive effect without violating a registrant's constitutional ex post facto protections. In so doing, MCL 8.5 provides two important guiding factors: (1) the remaining application must be consistent with the manifest intent of the Legislature; and (2) the remaining application must be operable, i.e., "otherwise complete in itself and capable of being carried out without reference to the unconstitutional sentence or provision," Rohan v. Detroit Racing Ass'n , 314 Mich. 326, 357, 22 N.W.2d 433 (1946). Under the Mendoza-Martinez punitive-effects analysis, this Court analyzed the aggregate effects of the 2011 SORA rather than the effects of each individual amendment. It is apparent, however, that a majority of the former SORA provisions underlying our conclusion that the 2011 SORA constitutes punishment were added by its 2006 and 2011 amendments. See, e.g., MCL 28.733 to MCL 28.736, as amended by 2005 PA 121 (creating the student-safety zones excluding registrants from living, working, or loitering within 1,000 feet of a school); MCL 28.727, as amended by 2011 PA 18, and MCL 28.725(1), as amended by 2011 PA 17 (adding events triggering an in-person reporting requirement and decreasing the reporting period to three days); MCL 28.722, as amended by 2011 PA 17 (creating the tiered classification system and basing SORA's requirements on those tiers). But even assuming that the retroactive application of the 2011 SORA without the specific provisions added by the 2006 and 2011 amendments would not violate constitutional ex post facto protections, the 2006 and 2011 amendments, in whole, cannot be excised from retroactive application because doing so renders the statute unworkable. The 2011 amendments completely restructured SORA through the imposition of a tiered classification system, and the duties and requirements of each registrant were based on that registrant's tier classification. Removing the 2011 amendments from SORA would render unclear who was required to comply with the act, see MCL 28.722(k), as amended by 2011 PA 17 (defining "listed offense" as a "tier I, tier II, or tier III offense"); how long each registrant must comply, see MCL 28.725(10) to (12), as amended by 2011 PA 17; how many times annually each registrant must report to law enforcement, see MCL 28.725a(3), as amended by 2011 PA 17; and what a registrant must show to petition for removal from registration, see MCL 28.728c, as amended by 2011 PA 18.

Outside the tiered classification system, certain discrete provisions of the 2006 and 2011 amendments—including the student-safety zones of MCL 28.733 to MCL 28.736, as amended by 2005 PA 121, and the in-person reporting requirements of MCL 28.725(1), as amended by 2011 PA 17—could be excised from retroactive application without affecting the statute's workability. However, even if the retroactive application of SORA without these discrete provisions were constitutional, that application would require this Court to engage in essentially legislative choices. Should this Court remove all in-person reporting requirements or only those beyond residence and employment changes? Should this Court retain the in-person reporting requirements but remove the "immediate" timeliness requirement? Or retain the reporting requirements and their timeliness requirement but remove the "in-person" requirement? Should this Court remove the student-safety zones completely or narrow their applicability to certain registrants who present a particular risk to children? Or narrow the student-safety zones to only "residing" within 1,000 feet of a school, removing the restrictions on employment and "loitering"? We decline to encroach on the Legislature's plenary authority to create law or on its role in shaping and articulating policy by choosing among the plethora of possibilities. See Ayotte v. Planned Parenthood of Northern New England , 546 U.S. 320, 329-330, 126 S. Ct. 961, 163 L. Ed. 2d 812 (2006) (characterizing the rewriting of state law as "quintessentially legislative work" and opining that crafting a remedy "where line-drawing is inherently complex may call for a far more serious invasion of the legislative domain than we ought to undertake") (quotation marks, citation, and punctuation omitted); People v. Steanhouse , 500 Mich. 453, 483-484, 902 N.W.2d 327 (2017) ( LARSEN , J., concurring) (stating that the Legislature "is certainly better equipped than this Court to weigh the policy options").

The prosecutor suggests that severance would not constitute problematic guesswork of legislative intent here because the Legislature has demonstrated its intent regarding the continued viability of SORA through its recent passage of 2020 PA 295. In light of the federal courts’ rulings in Does I and Does II that the 2011 SORA violates federal constitutional ex post facto protections, the Legislature chose to amend SORA to cure its constitutional infirmity. These amendments included the removal of the student-safety zones; the removal of the retrospective application of in-person reporting requirements for vehicle information, electronic mail addresses, Internet identifiers, and telephone numbers, MCL 28.725(2)(a) ; and the removal of registrants’ tier-classification information from the public website, MCL 28.728(3)(e). Considering that the Legislature removed these provisions from SORA through 2020 PA 295, it is argued that this Court's severance of similar provisions in the 2011 SORA from retroactive application would be consistent with the Legislature's intent and not constitute an unwise invasion into the legislative domain. We decline to adopt this proposed remedy. To begin, the intent of a prior legislature cannot be determined by looking at the actions of a subsequent one. See United States v. Price , 361 U.S. 304, 313, 80 S. Ct. 326, 4 L. Ed. 2d 334 (1960) ("[T]he views of a subsequent Congress form a hazardous basis for inferring the intent of an earlier one."); Nawrocki v. Macomb Co. Rd. Comm. , 463 Mich. 143, 177 n. 33, 615 N.W.2d 702 (2000). The 2019–2020 Legislature acted nearly 10 years after the 2011–2012 Legislature and did not consist of the same membership. Moreover, the 2019–2020 Legislature was considering not only a successful federal ex post facto challenge against SORA but also successful due-process and First Amendment challenges against SORA. See Does II , 449 F. Supp. 3d at 737-738.

Further, while the 2019–2020 Legislature did remove the provisions detailed earlier in this opinion, it did so at the same time it also introduced a bevy of other changes. These changes include both additional ameliorative changes and more restrictive changes. It is not altogether clear whether the Legislature would still advocate for the removal of the three provisions identified earlier without the addition of the several other amendments that were introduced in 2020 PA 295. Were we to sever the three specific provisions identified earlier, the resultant statute would be neither what the 2011–2012 Legislature intended with the creation of the 2011 SORA nor what the 2019–2020 Legislature intended with the enactment of 2020 PA 295. Accordingly, we do not believe that the passage of 2020 PA 295 supports the prosecution's proposed remedy for severing the 2011 SORA.

We also reject the proposal of amicus the Gratiot County Prosecutor's Office to remedy the constitutional violation by excising the particular provisions of the 2011 SORA that extend beyond its federal counterpart, the Sex Offender Registration and Notification Act (SORNA), 34 USC 20901 et seq. Specifically, amicus suggests that this Court excise from retroactive application the student-safety zones, MCL 28.733 to MCL 28.736, as amended by 2005 PA 121; the inclusion of the registrant's tier status on the public database, MCL 28.728(2)(l ), as amended by 2011 PA 18; and the in-person reporting requirements regarding temporary residences, the establishment of "any electronic mail or instant message address, or any other designations used in internet communications or postings," and the operation of vehicles, MCL 28.725(1)(e) through (g), as amended by 2011 PA 17. Amicus argues that because the Legislature's 2011 SORA amendments were intended to bring SORA into compliance with SORNA to avoid a reduction in federal funding, reforming SORA to match SORNA would be consistent with the Legislature's intent. And, because ex post facto challenges to SORNA have been rejected by the federal circuit courts, amicus argues that the constitutional error presented here would be cured. See, e.g., United States v. Parks , 698 F.3d 1, 5-6 (C.A. 1, 2012) ; United States v. Young , 585 F.3d 199, 204-206 (C.A. 5, 2009) ; United States v. Felts , 674 F.3d 599, 605-606 (C.A. 6, 2012) ; United States v. W.B.H. , 664 F.3d 848, 855-860 (C.A. 11, 2011).

Amicus is correct that legislative bill analyses regarding the 2011 SORA amendments indicate that the amendments would conform SORA to SORNA. See House Legislative Analysis, SB 188-189, 206 (March 22, 2011) (stating that the senate bills at issue would revise SORA "to conform to mandates under" SORNA and remarking that "[f]ailure to comply with SORNA will result in a state losing 10 percent of Byrne Justice Grant funding used to support law enforcement efforts"). But we have in the past been skeptical of the value of bill analyses in determining the Legislature's intent. See Frank W. Lynch & Co. v. Flex Technologies, Inc. , 463 Mich. 578, 587 n. 7, 624 N.W.2d 180 (2001) ("The problem with relying on bill analyses is that they do not necessarily represent the views of even a single legislator. Rather, they are prepared by House and Senate staff."). Further, the fact that the 2011 Legislature did not amend SORA to create an identical statutory scheme to SORNA and instead included several additional provisions indicates that the Legislature was, at the very least, not motivated solely by a desire to conform to SORNA. This proposed remedy raises again the prospect of this Court engaging in lawmaking on tenuous assumptions of the Legislature's intent, and we decline to do so.

Finally, in the absence of a remedy through severability, the prosecutor proposes that a former version of SORA can be applied to defendant through revival. In its usual application, revival occurs when an amendment of a statute is repealed and the former version of the statute is revived by the repeal of the amendatory provision. See Dykstra v. Holden , 151 Mich. 289, 293, 115 N.W. 74 (1908). Revival also applies when, instead of a legislative repeal of a statutory amendment, the courts find the amendment unconstitutional. When the amendment is constitutionally invalid, the statute behaves as if the amendment never existed. See, e.g., People v. Smith , 246 Mich. 393, 398, 224 N.W. 402 (1929) ("We must hold the amendment ... unconstitutional, and therefore no amendment. This holding leaves the law as it was before the abortive attempt to amend."); McClellan v. Stein , 229 Mich. 203, 213, 201 N.W. 209 (1924) ("We are therefore constrained to hold the law invalid, which leaves all preceding laws upon that subject in force."). Michigan has a legislative preference against revival, MCL 8.4, but it refers only to the legislative context of revival wherein the Legislature has acted to repeal an amendatory provision, not necessarily to the context wherein the courts have struck a provision down as unconstitutional.

Revival presents special challenges in the context of an ex post facto challenge to a statute with as complicated a legislative history as SORA. Our holding does not affect the prospective application of the 2011 SORA to registrants who committed listed offenses after 2011, from the time of their conviction to the effective date of the 2020 SORA amendments. Accordingly, it is not accurate to say that the SORA amendments failed to alter the statutory scheme, leaving the previous version in place unchanged, as with the usual revival context. Compare with Smith , 246 Mich. at 398, 224 N.W. 402 ; McClellan , 229 Mich. at 213, 201 N.W. 209. It is possible that revival could nonetheless be applied only to pre-2011 registrants under a theory that the amendments were invalid as to retroactive application only, leaving previous SORA formulations active. However, doing so raises the same concerns of legislative infringement and practical complications discussed in conjunction with severability, and the prosecutor has offered no response to these concerns raised by defendant and amici.

The Legislature has modified SORA over the past nearly 30 years in a series of amendments introducing new provisions; contracting, expanding, and removing established provisions; creating new ameliorative provisions; and in the case of the 2011 amendments, completely restructuring the statutory scheme. Accordingly, SORA presents a different situation altogether than the prototypical single statutory amendment that represents the Legislature's intent to change a singular provision of the law and that can be neatly foreclosed from certain applications. In this case, given the extensive legislative history of SORA, it is unclear whether revival of earlier SORA formulations is consistent with the Legislature's intent. See Ayotte , 546 U.S. at 330, 126 S.Ct. 961 ("[T]he touchstone for any decision about remedy is legislative intent, for a court cannot use its remedial powers to circumvent the intent of the legislature.") (quotation marks and citation omitted). And, although not a dispositive obstacle, the sheer volume of past versions of SORA poses significant administrative difficulties for the Michigan State Police in attempting to define and enforce multiple different registry schemes and for registrants in attempting to adhere to the requirements of an out-of-date registry scheme.

Having determined that severability and revival are inappropriate tools to remedy the constitutional violation in this case, we are constrained to hold that the 2011 SORA may not be retroactively applied to registrants whose criminal acts subjecting them to registration occurred before the enactment of the 2011 SORA amendments.

VI. CONCLUSION

We hold that the 2011 SORA, when applied to registrants whose criminal acts predated the enactment of the 2011 SORA amendments, violates the constitutional prohibition on ex post facto laws. As applied to defendant Betts, because the crime subjecting him to registration occurred in 1993, we order that his instant conviction of failure to register as a sex offender be vacated.

The case is remanded for further proceedings consistent with this opinion.

McCormack, C.J., and Bernstein and Cavanagh, JJ., concurred with Clement, J.

Zahra, J. (concurring in part and dissenting in part).

I join Parts I and II of Justice VIVIANO ’s opinion concurring in part and dissenting in part. I agree with his application of this Court's severability precedents to Michigan's Sex Offenders Registration Act, MCL 28.721 et seq. , as amended by 2011 PA 17 and 18. But I decline to join Part III because it is unnecessary to the resolution of this case.

Viviano, J. (concurring in part and dissenting in part).

I generally agree with the majority's holding that Michigan's Sex Offenders Registration Act (SORA), MCL 28.721 et seq. , as amended by 2011 PA 17 and 18 (the 2011 SORA), violates the Ex Post Facto Clauses of the state and federal Constitutions. But I disagree that the statute is not severable and would conclude that the unconstitutional portions of the statute can be removed to the extent necessary in this case. But doing so requires greater care than the majority offers in specifying the constitutional infirmities in the statute. The difficulties presented by the majority's analysis, as well as other problems posed by current precedent, should also lead us to consider whether a different approach to the severability analysis is needed and whether that approach better reflects the requirements of MCL 8.5, the general statute on severability.

I. PRINCIPLES OF SEVERABILITY

Our "Court has long recognized" that unconstitutional portions of a statute are not to be given effect if the constitutional portions of the statute remain operable. See In re Request for Advisory Opinion Regarding Constitutionality of 2011 PA 38 , 490 Mich. 295, 345, 806 N.W.2d 683 (2011). Our general rule therefore favors severability. Id. ; see also 2 Singer, Sutherland Statutes and Statutory Construction (7th ed., November 2020 update), § 44:1 ("There is a presumption in favor of severability."). As Justice Thomas Cooley wrote, "It would be inconsistent with all just principles of constitutional law to adjudge these enactments void, because they are associated in the same act, but not connected with or dependent on others which are unconstitutional." Cooley, Constitutional Limitations (1868), p. 177. The partial unconstitutionality of a statute "does not authorize the courts to declare the remainder void also," unless the provisions are so entwined that the remaining portion is not "complete in itself, and capable of being executed wholly independent of that which was rejected ...." Cooley, p. 178.

The Legislature has codified this favorable view of severability in MCL 8.5 :

In the construction of the statutes of this state the following rules shall be observed, unless such construction would be inconsistent with the manifest intent of the legislature, that is to say:

If any portion of an act or the application thereof to any person or circumstances shall be found to be invalid by a court, such invalidity shall not affect the remaining portions or applications of the act which can be given effect without the invalid portion or application, provided such

remaining portions are not determined by the court to be inoperable, and to this end acts are declared to be severable.^{[ ]}

Under this statute, a court must "consider, first, whether the Legislature expressed that the provisions at issue were not to be severed from the remainder of the act." Blank v. Dep't of Corrections , 462 Mich. 103, 123, 611 N.W.2d 530 (2000). "If it did not, then [a court] must determine whether the unconstitutional portions are so entangled with the others that they cannot be removed without adversely affecting the operation of the act." Id. Put differently, a court must determine that (1) the Legislature manifested an intent to remove the statute at issue from the general presumption of severability in MCL 8.5, and if not (2) whether the act is "capable of separation in fact" in that the constitutional portions represent an operable whole. 2 Singer, Sutherland Statutes and Statutory Construction (7th ed., November 2020 update), § 44:3. In resolving the first question, we have considered whether "the Legislature ‘would have passed the statute had it been aware that portions therein would be declared to be invalid and, consequently, excised from the act.’ " In re Request for Advisory Opinion , 490 Mich. at 346, 806 N.W.2d 683 (citation omitted). II. ANALYSIS

A. THE LEGISLATURE'S MANIFEST INTENT

The first question when applying MCL 8.5 is whether the Legislature manifested its intent to make SORA inseverable. The majority fails to point to, nor have I discovered, any statutory text or other material that would suggest—let alone "manifest"—the Legislature's intent to shield SORA from the normal presumption favoring severability. Nor does the majority address whether the Legislature would have enacted SORA without its unconstitutional portions. As discussed below, the nature of this question focuses on the Legislature's unexpressed intentions and thus is difficult to answer. Nevertheless, under my analysis below, only discrete portions of the statute would be severed. I have a difficult time believing that the Legislature, had it reflected on this possibility, would prefer that a complex and far-reaching statute like SORA should be eliminated simply because a few insular sections have been removed. Accordingly, there is no basis for concluding that the Legislature has manifested an intent that SORA be inseverable.

B. OPERABILITY OF THE VALID PORTIONS

The central question therefore is whether severing the unconstitutional portions of SORA leaves a complete and operable statute in place. The majority proclaims that even if removing the 2006 and 2011 amendments from SORA resulted in a constitutional statute, those amendments "cannot be excised from retroactive application because doing so renders the statute unworkable." But severance does not require taking a machete to the statute—few statutes would remain operable after that approach. Instead, "[w]hen confronting a constitutional flaw in a statute, the court should try to invalidate no more of the statute than necessary." 2 Singer, Sutherland Statutes and Statutory Construction (7th ed., November 2020 update), § 44:4; see also Alaska Airlines, Inc. v. Brock , 480 U.S. 678, 684, 107 S. Ct. 1476, 94 L. Ed. 2d 661 (1987) (" ‘[A] court should refrain from invalidating more of the statute than is necessary ....’ ") (citation omitted).

When considering whether smaller portions could be severed, the majority acknowledges that two pieces of the statute—the student-safety zones in MCL 28.733 to 28.736, as amended by 2005 PA 121, and the in-person reporting requirements in MCL 28.725(1), as amended by 2011 PA 17—"could be excised from retroactive application without affecting the statute's workability." In other words, the majority essentially admits that we could sever two portions of the statute and leave the rest operable. Of course, finding that the rest of the statute could remain operable without these requirements is not difficult—SORA did, in fact, operate without them before the 2006 and 2011 amendments that added them.

Under these circumstances, MCL 8.5 requires severance. And yet the majority shies away from this conclusion because deciding which parts to sever, in these circumstances, involves "essentially legislative choices." But the relevant legislative choice here was made by the Legislature when it enacted MCL 8.5. And it is hard to see how MCL 8.5 could survive the majority's logic; if the decision on how to sever certain "discrete" portions of SORA is impermissibly legislative, then severability would never be permissible. See Fallon, Facial Challenges, Saving Constructions, and Statutory Severability , 99 Tex. L. Rev. 215, 224 (2020) ("[C]haracterizations of the judicial role in severing statutes as involving an impermissible ‘rewriting’ prove too much insofar as they imply that courts should never sever statutes with invalid applications that Congress sought to prescribe."). By rejecting the Legislature's choice, codified in MCL 8.5, the majority reaches the baffling conclusion that wiping out an entire statute is more respectful of legislative intent than removing a few words or sections.

C. APPLICATION

Severing the unconstitutional portions of the statute does not require legislative decision-making. It does, however, require precision in defining the unconstitutional sections. The majority assesses the "aggregate effects of the 2011 SORA rather than the effects of each individual amendment." While the United States Supreme Court suggested such an analysis in Smith v. Doe , 538 U.S. 84, 99-100, 123 S. Ct. 1140, 155 L. Ed. 2d 164 (2003), neither that Court nor ours has extended this mode of analysis to the question of severability. The United States Supreme Court itself has stated that portions of a statute that violate the Ex Post Facto Clause might be severed. See Weaver v. Graham , 450 U.S. 24, 36 n. 22, 101 S. Ct. 960, 67 L. Ed. 2d 17 (1981).

The arguments in this case have generally focused on the in-person reporting requirements, the student-safety zones, and the public notification of the tiered-classification system. My analysis likewise centers on these provisions. For the reasons that follow, I would sever a few words from the reporting requirement, I would not decide how much or little to sever of the student-safety zones, and I would not sever any of the tiered-classification system, which I do not believe is unconstitutional. 1. IMMEDIATE IN-PERSON REPORTING REQUIREMENT

At the time of defendant's present conviction for violating SORA, this requirement provided that an individual who resides in Michigan and is required to register under SORA "shall report in person and notify the registering authority having jurisdiction where his or her residence or domicile is located immediately after" various events occur, including changes of residence or domicile, establishment of e-mail addresses or designations used on the Internet, and when he or she "purchases or begins to regularly operate any vehicle" or discontinues ownership or use of the vehicle. MCL 28.725(1)(a), (f), and (g), as amended by 2011 PA 17. "Immediately" is defined in the statute to mean "within 3 business days." MCL 28.722(g), as amended by 2011 PA 17. These were the provisions defendant pleaded guilty to violating. In determining whether and how to sever this provision, we must go further than the majority in isolating the requirement's unconstitutional aspects under the relevant factors of the ex post facto analysis.

The majority's analysis demonstrates that this provision is unconstitutional punishment because it required immediate in-person reporting on a host of quotidian events, such as signing up for a new e-mail account. As defendant has persuasively argued, the need to immediately report in person is what restrains and disables him, which is one of the factors in the ex post facto analysis applicable here. See Smith , 538 U.S. at 99-100, 123 S.Ct. 1140. Standing alone, a reporting requirement is not disruptive or restraining. Forcing a registrant to call or otherwise contact the authorities, even "immediately," i.e., within three days, is not overly burdensome. But the requirement that the registrant arrange their affairs so that they can show up in person within three days after relatively routine events is, as the majority observes, a significant burden.

With regard to the excessiveness of the requirements in relation to a nonpunitive purpose, it is again the need to report in person within three days that proves problematic. The desire to keep close tabs on registrants by requiring frequent reporting bears a reasonable relationship to the nonpunitive purpose of protecting the public. The majority and various parties and amici cite statistical research indicating that sex offenders do not have unusually high recidivism rates. However, I am not yet ready to say that the Legislature was unreasonable in requiring frequent reporting to combat recidivism. For one thing, the research rests on data concerning sex offenders who were caught committing a subsequent offense. See, e.g., Hanson et al., High-Risk Sex Offenders May Not Be High Risk Forever , 29 J Interpersonal Violence 2792, 2796 (2014) (defining "offense-free" as "no new sexual offenses were detected during [the] time period"). And it is well established that sex crimes are seriously underreported. See Morgan & Kena, U.S. Department of Justice, Criminal Victimization, 2016: Revised (October 2018, NCJ 252121), p. 7 (showing that in 2016 only 23.2% of rapes and sexual assaults were reported, making it the most underreported class of crimes). As a result, it remains possible, if not likely, that the recidivism rates reported in the studies " ‘underestimate the risk an offender will commit an offense over [his or her] lifetime.’ " Belleau v. Wall , 811 F.3d 929, 933 (C.A. 7, 2016), quoting DeClue & Zavodny, Forensic Use of the Static-99R: Part 4. Risk Communication , 1 J. Threat Assessment & Mgmt. 145, 149 (2014); Scurich & John, Abstract, The Dark Figure of Sexual Recidivism , 37 Behav. Sci. & L. 158 (2019) ("Virtually all of the studies [of sexual offender recidivism] define recidivism as a new legal charge or conviction for a sexual crime .... It is uncontroversial that such a definition of recidivism underestimates the true rate of sexual recidivism because most sexual crime is not reported to legal authorities, a principle known as the ‘dark figure of crime.’ ... Under any configuration of assumptions, the dark figure is substantial, and as a consequence the disparity between recidivism defined as a new legal charge or conviction for a sex crime and recidivism defined as actually committing a new sexual crime is large. These findings call into question the utility of recidivism studies that rely exclusively on official crime statistics ...."). Even were I more inclined to credit the studies on which the majority relied, I would defer to the Legislature on such matters when there is room for debate. Given the nature of our role of adjudicating individual disputes and the consequent institutional limitations this role entails, we must exercise "humility about the capacity of judges to evaluate the soundness of scientific and economic claims[.]" Barrett, Countering the Majoritarian Difficulty , 32 Const. Comment 61, 74 (2017) (reviewing Barnett, Our Republican Constitution: Securing the Liberty and Sovereignty of We the People (New York: HarperCollins, 2016)).

Thus, the immediate reporting requirements are not excessive standing alone. What makes them excessive is the need to report in person. There has been no evidence put forward to believe that registrants are particularly apt to shirk their reporting obligation or to make false reports if they are not in person. Indeed, it is hard to see any connection between the in-person requirement and the contents of the required reports. For example, how does showing up in person make it more or less believable that the registrant really changed his or her e-mail address? I would therefore find that the in-person requirement is what transforms the immediate reporting requirement into a prohibited punishment under the Ex Post Facto Clauses.

Having pinpointed the source of the constitutional infirmity, the severance analysis is straightforward: I would sever the phrase "report in person and" from MCL 28.725(1), as amended by 2011 PA 17. The statute will still require that the registrant "notify the registering authority" when the various triggering events occur. In this respect, the statute would resemble its pre-2011 version, which similarly required the registrant to "notify" the appropriate authorities. MCL 28.725(1), as amended by 2006 PA 402. The reporting requirement would therefore remain valid and operable, as would the remainder of SORA; only the need to make the reports in person would be removed from the statute.

2. STUDENT-SAFETY ZONES

With regard to the student-safety zones, the various questions posed by the majority about what to sever are largely misplaced. Even if the sections creating these zones were struck down in their entirety, the majority admits that the remaining provisions of SORA would be operable. These sections are tucked in a separate corner of SORA called "article II." MCL 28.733 through MCL 28.736, as amended by 2005 PA 121. The zones are a discrete requirement that does not involve registration itself but rather a distinct limitation imposed on registrants. They are not entwined with the rest of SORA—they refer to SORA only to note that the geographic restrictions apply to individuals required to register under SORA "article II," MCL 28.723 through MCL 28.730.

There is no need in this case to decide which parts of these sections should be severed. Defendant was not convicted under these provisions. Indeed, even he admits that if they were severed, his conviction must be upheld. All that MCL 8.5 requires is that the remaining constitutional portions of the statute be operable. Here, as noted, under any conceivable severance of the student-safety zones, the remaining portions would be operable, and the present dispute—whether defendant's conviction can be upheld—would be resolved. In other words, a decision on how to sever the student-safety zones has no relevance in resolving the case before the Court. It is, therefore, rather astounding that the majority would strike down the entire statute because, among other things, deciding how to do something that does not need to be done—sever the student-safety zones—would be an impermissible legislative action. See People v. McMurchy , 249 Mich. 147, 160, 228 N.W. 723 (1930) ("In Brazee v. Michigan , [241 U.S. 340, 36 S. Ct. 561, 60 L. Ed. 1034 (1916),] the court held that it was not necessary to go into the constitutionality of certain clauses of an act, where the act was severable and defendant had been convicted under a part of the act, the constitutionality of which could not be questioned."). In sum, then, I agree that the student-safety zones are unconstitutional but would not decide in this case whether they could be severed in a manner that renders the remaining portions of these sections constitutional.

3. TIERED-CLASSIFICATION SYSTEM

The tiered-classification system is a different story. Under this framework, the registrant is publicly placed into one of three tiers depending on the offense of which he or she was convicted. The majority does not spend much time explaining the constitutional infirmities with the classification system. It notes that SORA's public broadcasting of information resembles the historic punishment of shaming. And it observes that SORA resembles the aim of retribution because it classifies individuals without regard to individualized risk assessments. But once again, it is unclear which specific provisions the majority finds constitutionally troublesome for purposes of the severability analysis. As amicus the Gratiot County Prosecutor rightly notes, however, the crux of defendant's argument was not with the lack of individualized risk assessments (although he does cover that) but more specifically with the public nature of the tiered classifications.

Specificity matters with respect to the tiered system. The majority correctly explains that the 2011 amendments "restructured SORA through the imposition of a tiered classification system, and the duties and requirements of each registrant were based on that registrant's tier classification." Severing the tiers would, as the majority concludes, undoubtedly result in an unworkable statute. The issues therefore are whether and to what extent the tiered system is unconstitutional and can be severed. There are two aspects of this issue that must be addressed: (1) Is it punitive to use criminal offenses as the basis for the tiered classification rather than an individualized risk assessment? and (2) Does the public availability of the registrant's tier classification constitute punishment?

With regard to individualized risk assessments, I struggle to see how the Legislature is imposing a punishment by tying registration classifications to the offense of which the individual was convicted. See MCL 28.722(k) and MCL 28.722(s) through (u), as amended by 2011 PA 17. The Legislature can reasonably conclude that violation of certain crimes portends a greater risk of recidivism or danger to the public than does violation of other crimes, and it can adjust the regulatory requirements accordingly. As the United States Supreme Court noted:

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.... As stated in Hawker [v. New York , 170 U.S. 189, 197, 18 S. Ct. 573, 42 L. Ed. 1002 (1898) ]: "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. [ Smith , 538 U.S. at 103-104, 123 S.Ct. 1140.]

Given the Supreme Court's holding, it is not surprising that the federal sex offender registration statute similarly links the registrant's tier classification to the type of offense of which he or she was convicted. See 34 USC 20911 .

Moreover, it is not clear that an individualized risk assessment offers a superior means for accurately appraising the probability that the registrant will commit another sex offense. An assessment tool like the Static-99R that the Attorney General endorses here produces "estimates ... [that] pertain only to the odds that the released offender will subsequently be arrested for or convicted of—in short, detected—committing further sex crimes." Belleau , 811 F.3d at 933. As noted above, the data used in such an assessment relates to the risk of detection rather than the risk that the registrant will actually commit a new offense, whether detected or not; as a result, it might underestimate the relevant probability. Id. For these reasons, I cannot see how the lack of such a metric and the reliance on the convicted offense constitutes a punishment. Furthermore, the publication of the registrant's tier classification is not a punishment. Indeed, where, as here, the tier simply reflects the underlying offense, the tier classification itself provides the public with no new information. Cf. Smith , 538 U.S. at 101, 123 S.Ct. 1140 ("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.").

Therefore, I would not find the tiered-classification system to be unconstitutional.

D. SUMMARY

In short, while I generally concur in the majority's conclusion that SORA violates the Ex Post Facto Clauses of the state and federal Constitutions, I disagree that the offending statutory provisions applicable here are inseverable. The severability analysis requires a more exacting appraisal of the constitutional problems with particular provisions. In undertaking that examination, I would sever the immediate in-person reporting requirement from MCL 28.725(1), as amended by 2011 PA 17. I would also hold that although the student-safety zones are unconstitutional, it is unnecessary to decide whether those provisions could be severed in a manner that allows any portion of them to remain. Finally, I would hold that the publicly available tiered-classification system based on the registrant's conviction is not unconstitutional.

My analysis would require upholding defendant's conviction, given that he violated the severed version of MCL 28.725(1), as amended by 2011 PA 17. As severed, the provision still required him to register and report certain information to the authorities. His failure to do so violated the valid portions of the statute.

III. HISTORICAL APPROACH TO SEVERABILITY

To the extent that the majority's opinion reflects existing precedent—specifically, the focus on the Legislature's hypothetical intentions—the majority's opinion raises questions that should be considered in an appropriate future case. In particular, I would consider whether our precedent has focused too heavily on legislative intent and whether a more historically grounded approach to severability would better reflect the nature of judicial decision-making and the text of MCL 8.5.

The historical approach to severability rests on a few fundamental principles. Our courts do not sit as councils of revision, wielding a pen to strike out the offending portions of the statute or to remove the law from the statute books. See Citizens Protecting Michigan's Constitution v. Secretary of State , 503 Mich. 42, 92 n. 149, 921 N.W.2d 247 (2018) ("Despite our ruling [that an enacted provision is unconstitutional], we have no power to make the law disappear."). Our authority is limited to the exercise of judicial power, by which we can "hear and determine controversies between adverse parties, and questions in litigation." Daniels v. People , 6 Mich. 381, 388 (1859). The judicial " ‘power exercised is that of ascertaining and declaring the law applicable to the controversy.’ " Seila Law LLC v. Consumer Fin. Protection Bureau , 591 U.S. ––––, ––––, 140 S. Ct. 2183, 2219, 207 L. Ed. 2d 494 (2020) (Thomas, J., concurring in part and dissenting in part), quoting Massachusetts v. Mellon , 262 U.S. 447, 488, 43 S. Ct. 597, 67 L. Ed. 1078 (1923). "In the context of a constitutional challenge, ‘[i]t amounts to little more than the negative power to disregard an unconstitutional enactment.’ " Seila Law LLC , 591 U.S. at ––––, 140 S. Ct. at 2219 (Thomas, J., concurring in part and dissenting in part), quoting Mellon , 262 U.S. at 488, 43 S.Ct. 597. Given the nature of our power, we cannot "excise, erase, alter, or otherwise strike down a statute." Seila Law LLC , 591 U.S. at ––––, 140 S. Ct. at 2220 (Thomas, J., concurring in part and dissenting in part).

In light of these central principles, courts historically did not claim to sever or strike down statutory language when facing statutes that were partially unconstitutional. Instead, they would simply apply the challenged statute together with the Constitution to the case at hand; if the statute conflicted with the Constitution, courts held the "law void to the extent of repugnancy," but "there was no ‘next step’ in which courts inquired into whether the legislature would have preferred no law at all to the constitutional remainder" of the statute. Walsh, Partial Unconstitutionality , 85 NYU L. Rev. 738, 777 (2010) ; see also Murphy v. Nat'l Collegiate Athletic Ass'n , 584 U.S. ––––, ––––, 138 S. Ct. 1461, 1486, 200 L. Ed. 2d 854 (2018) ; cf. Tribe, The Legislative Veto Decision: A Law by Any Other Name? , 21 Harv. J. on Legis. 1, 25-26 (1984) (advocating for a similar approach, under which courts are not regarded "as choosing how much or how little of a law to ‘strike down’ but as resolving controversies in a manner that rejects only such claims based upon a given law as are themselves deemed incompatible with the Constitution"). The general question, in other words, was whether the portion of the statute at issue could be applied in the case at hand and not whether the unconstitutional parts of the statute, unrelated to the case, precluded enforcing any part of the statute. In essence, the Constitution was found to have displaced the statute to the extent the statute contravened the Constitution. Partial Unconstitutionality , 85 NYU L. Rev. at 742 ("[U]nder a displacement-based approach, a court does not excise anything from a statute but instead determines the extent to which superior law displaces inferior law in resolving the particular case before it."). Of course, some constitutional defects might infect the rest of the statute to such an extent that no operable portions remain. See The Legislative Veto Decision , 21 Harv J. on Legis. at 25.

The historical approach would appear to solve some of the problems with the current framework. One of the most significant difficulties is with the proposition that severability requires a court to determine whether the Legislature would have passed the statute without the unconstitutional portions had it known of their defects. See In re Request for Advisory Opinion , 490 Mich. at 345, 806 N.W.2d 683. This question essentially forces a court to speculate about what the Legislature intended should occur if a statute is found partially unconstitutional—yet, the Legislature likely never thought about that scenario and did not provide for it through enacted text. The answer to the question—to the extent there is one—will be difficult to ascertain, and the search for it will take courts away from their prescribed role in determining what the statutory text means. See Murphy , 584 U.S. at ––––, 138 S. Ct. at 1486-1487 (Thomas, J., concurring). As Justice Thomas wrote, the modern approach "requires judges to determine what Congress would have intended had it known that part of its statute was unconstitutional. But it seems unlikely that the enacting Congress had any intent on this question[.]" Id. at ––––, 138 S. Ct. at 1486-1487. And, critically, "intentions do not count unless they are enshrined in a text that makes it through the constitutional process of bicameralism and presentment." Id. at ––––, 138 S. Ct. at 1487. See also Note, Constitutional Avoidance, Severability, and a New Erie Moment , 42 Harv. J. L. & Pub. Pol'y 649, 650 (2019) (criticizing the assumption, sometimes relied on in severability analyses, of "the existence of an unexpressed legislative intent that judges can discover").

Another potential problem with the modern approach is that it enables a court to pass on the constitutionality of provisions that have scarce relationship to the case before the court. In other words, it potentially enables parties to challenge statutory provisions that they might lack standing to challenge. See Murphy , 584 U.S. at ––––, 138 S. Ct. at 1487 ("If one provision of a statute is deemed unconstitutional, the severability doctrine places every other provision at risk of being declared nonseverable and thus inoperative" irrespective of whether the plaintiff had standing to attack those provisions.). Severability might enable parties to evade a constitutionally valid statutory provision that applies to the dispute simply because other parts of the statute, which do not apply in the case, are unconstitutional. See Zimmerman, Supplemental Standing For Severability , 109 Nw. U. L. Rev. 285, 304 (2015) (noting the possibility that a party could "argue that, even if the part of the statute that applies to them is constitutional, that part is invalid because some other part of the statute is unconstitutional and cannot be severed"); see generally 2 Singer, Sutherland Statutes and Statutory Construction (7th ed., November 2020 update), § 44:2 (noting that severability raises this possibility "whenever a person not subject to the invalid provision, but nevertheless within the scope of the statute, seeks to attack the act by showing the entire act to be invalid by reason of the invalidity of a part"). A few federal courts have found that a party lacks standing to make such arguments, although the United States Supreme Court has addressed such arguments without questioning standing. See Supplemental Standing , 109 Nw. U. L. Rev. at 306-307 (discussing the caselaw). In an appropriate future case, I would consider whether our precedent has gone off track with its focus on legislative intent and the need to address parts of statutes inapplicable to the case at hand. I would also consider whether MCL 8.5 is consonant with the historical approach and thereby avoids the possible problems discussed above. In particular, the Court should examine whether MCL 8.5 allows courts to enforce any valid provisions that can stand alone regardless of the constitutionality of the statute's other provisions. Or does MCL 8.5 require courts to examine the entire statute at issue and pencil off the portions that are unconstitutional? IV. CONCLUSION

My questions concerning the historical approach to severability are for another day. Applying the plain language of MCL 8.5 in light of current precedent, I would conclude that the unconstitutional portions of SORA are severable and that defendant's conviction must be upheld. I therefore dissent from the majority's contrary conclusions.

Zahra, J. (except as to Part III), concurred with Viviano, J.

Welch, J., did not participate in the disposition of this case because the Court considered it before she assumed office.