People v. Kardasz: Michigan Supreme Court Holds 2021 SORA Is Punishment But Not “Cruel or Unusual”

Court: Michigan Supreme Court
Case: People of the State of Michigan v. Robert James Kardasz, No. 165008 (Dec. 19, 2025)
Author of Majority: Justice Thomas (joined by Cavanagh, Welch, Bolden, Hood)
Separate Opinions: Justice Zahra (concurring in part/dissenting in part); Justice Bernstein (concurring in part/dissenting in part)

I. Introduction

People v. Kardasz is a major decision in Michigan constitutional law and in the law of sex offender registration. The Michigan Supreme Court holds that the 2021 version of the Sex Offenders Registration Act (SORA), as amended by 2020 PA 295, constitutes “punishment” for constitutional purposes, but that it is not “cruel or unusual punishment” under Article 1, § 16 of the 1963 Michigan Constitution.

The ruling:

Affirms that the 2021 SORA is a punitive regime, not a purely civil regulatory measure.
Holds that, even as punishment, the 2021 SORA is neither facially nor as applied to the defendant “cruel or unusual” under the Michigan Constitution.
Declines to conduct an independent Eighth Amendment analysis, reasoning that Michigan’s “cruel or unusual” clause is more protective and the statute passes under that higher standard.
Denies leave to appeal on all issues related to lifetime electronic monitoring (LEM), leaving prior Court of Appeals precedent and out-of-state authority on LEM largely undisturbed.

In doing so, Kardasz:

Builds on and clarifies earlier Michigan precedents (Betts and Lymon) that also treated modern SORA schemes as punitive.
Parts ways with the U.S. Supreme Court’s more deferential reading of Alaska’s earlier, less-intrusive regime in Smith v Doe.
Signals to the Legislature that key structural features of SORA—offense-based tiering, lifetime registration without meaningful exit, and broad public web publication—are “excessive” relative to their stated public-safety goals, even though they are not (yet) constitutionally forbidden.

II. Background and Procedural History

A. The Underlying Crime and Sentences

Robert James Kardasz was convicted by a Macomb County jury of first-degree criminal sexual conduct (CSC-I) under MCL 750.520b(1)(a) and (2)(b) for the sexual assault of his five-year-old daughter. CSC-I is among the most serious sexual offenses in Michigan.

Initially, the trial court imposed a prison term above the statutory mandatory minimum of 25 years. On appeal, the Court of Appeals affirmed the conviction but vacated the sentence, remanding for resentencing because the trial court had exceeded the 25-year minimum without adequately explaining why a 30-year minimum was proportionate.

On remand, the trial court resentenced him to 25–40 years’ imprisonment (i.e., the statutory minimum range under MCL 750.520b(2)(b)). The court also:

Imposed lifetime electronic monitoring (LEM) under MCL 750.520n, and
Required lifetime registration under SORA.

B. Appeals and Issues Presented

In the second appeal, Kardasz challenged:

SORA, arguing that lifetime registration under the 2021 SORA is:
- Facially and as applied “cruel or unusual punishment” under Const 1963, art 1, § 16, and
- “Cruel and unusual punishment” under the Eighth Amendment.
LEM, arguing that:
- LEM is an unreasonable search under the Fourth Amendment and the Michigan analogue, and
- LEM itself is cruel or unusual (state) and cruel and unusual (federal) punishment.

The Court of Appeals rejected all of his challenges and affirmed. Kardasz obtained an order from the Michigan Supreme Court directing oral argument on his application for leave, focused on:

Whether lifetime SORA registration is cruel or unusual / cruel and unusual.
Whether lifetime LEM without individualized risk assessment or an exit mechanism is cruel or unusual / cruel and unusual.
Whether lifetime LEM is cruel or unusual as applied.
Whether lifetime LEM is an unreasonable search under state or federal constitutional provisions.

After oral argument and supplemental briefing (including the relationship between parole and LEM), the Court:

Reached the SORA issues on the merits, and
Denied leave with respect to all LEM-related issues (thus leaving LEM’s constitutionality largely unaddressed at the Supreme Court level in this case).

III. Summary of the Court’s Holdings

A. Majority (Justice Thomas)

2021 SORA constitutes “punishment” under the Michigan Constitution.
Although the Legislature labeled SORA as civil and preventive, the Court—applying the Mendoza‑Martinez/Smith v Doe factors as incorporated in Michigan law—concludes by the “clearest proof” standard that:
- SORA resembles historical punishments (parole and shaming) in important respects.
- SORA imposes significant affirmative restraints and disabilities.
- SORA promotes traditional aims of punishment, particularly retribution.
- While it is rationally connected to public safety, multiple design features are excessive in relation to that purpose.
Nonetheless, 2021 SORA is neither facially nor as applied “cruel or unusual” punishment under Const 1963, art 1, § 16.
Using the Lorentzen/Bullock proportionality framework (a multi-factor gross-disproportionality test), the Court holds:
- The severity of lifetime registration and periodic reporting is not disproportionate to the gravity of Tier III offenses like CSC-I.
- Compared to sentences for other serious Michigan offenses, tiered SORA burdens on Tier III offenders do not appear grossly out of line.
- In national context, Michigan’s scheme is stringent but not uniquely extreme, and no clear national trend renders it grossly excessive.
- SORA undermines rehabilitation and is problematic on that dimension, but that single factor does not outweigh the others.
Therefore:
- Facial challenge fails: SORA is not grossly disproportionate in all its applications.
- As-applied challenge fails: given the seriousness of defendant’s offense and the record, Tier III registration is not grossly disproportionate as applied to him.
No separate Eighth Amendment analysis is performed.
Because Article 1, § 16 is more protective than the federal Eighth Amendment, and SORA passes under the more protective state standard, the Court finds it unnecessary to analyze the federal claim independently.
Court of Appeals’ judgment is affirmed in result but partially vacated in reasoning.
The Supreme Court:
- Affirms the Court of Appeals’ bottom-line holding that 2021 SORA is not cruel or unusual punishment, but
- Vacates that opinion to the extent it is inconsistent with the Supreme Court’s reasoning (especially the holding that SORA is punishment).
Leave is denied on all LEM issues.
The Court expressly declines to decide:
- Whether lifetime LEM is an unreasonable search under state or federal law.
- Whether lifetime LEM is cruel or unusual, facially or as applied.

B. Separate Opinions

Justice Zahra (concurring in part, dissenting in part)
- Agrees that 2021 SORA is not cruel or unusual punishment.
- Disagrees that it is punishment at all; argues it is a valid civil regulatory scheme.
- Contends defendant has not met the “clearest proof” standard needed to override legislative intent; reads Smith v Doe as controlling and supportive of SORNA-type schemes as civil.
- Would affirm the Court of Appeals’ conclusion that SORA is nonpunitive and vacate its suggestion that SORA is punishment as applied to sexual offenses.
- Warns that the majority is eroding deference to legislative judgments and undermining public safety-related registries.
Justice Bernstein (concurring in part, dissenting in part)
- Agrees with the ultimate result: the 2021 SORA is not cruel or unusual punishment.
- However, he would not reach the punishment vs. civil question; he would simply “assume without deciding” that SORA is punishment and resolve the case on proportionality grounds alone.
- Dissents only from Part III(B) of the majority (the punishment analysis).

IV. Precedential Landscape and Authorities Cited

A. Prior Michigan Cases on SORA and Punishment

People v Betts, 507 Mich 527 (2021)
The Court held that the 2011 SORA amendments (including student safety zones, extensive in-person reporting, and public tier publication) were punitive and that their retroactive application violated state and federal ex post facto protections. Betts:
- Applied the Mendoza‑Martinez/Smith factors to SORA.
- Emphasized banishment-like geographic restrictions, heavy reporting burdens, and their tenuous relationship to actual risk reduction.
People v Lymon, ___ Mich ___ (2024)
Lymon addressed the 2021 SORA, but only as applied to offenders whose crimes had no sexual component (e.g., kidnapping/false imprisonment of children). The Court there:
- Held that the Legislature intended SORA to be civil.
- Nevertheless found that, as applied to nonsexual offenses, the 2021 SORA was punishment and cruel or unusual under the Michigan Constitution due to its disproportionality and misfit with its nonpunitive purpose.
- Invalidated SORA as applied to those nonsexual offenders, but left its application to sexual offenders undecided—an issue now resolved in Kardasz.
People v Earl, 495 Mich 33 (2014)
Provides Michigan’s basic two-step framework for distinguishing punishment from civil remedies:
1. Legislative intent (civil or criminal).
2. Whether, despite a civil label, the scheme is “so punitive in purpose or effect” as to override that intent (using Mendoza‑Martinez factors and “clearest proof” standard).
People v Bullock, 440 Mich 15 (1992) & People v Lorentzen, 387 Mich 167 (1972)
These cases establish Michigan’s gross disproportionality test under Article 1, § 16, now routinely applied to “cruel or unusual” claims:
- Severity vs. gravity of the offense.
- Comparison to penalties for other offenses in Michigan.
- Comparison to penalties for the same offense in other states.
- Consideration of rehabilitation and broader penological policies.
People v Parks, 510 Mich 225 (2022) & People v Taylor, ___ Mich ___ (2025)
Recent decisions invalidating mandatory life without parole for 18-year-olds (Parks) and extending similar analysis to 19- and 20-year-olds (Taylor). These cases are cited in Kardasz to:
- Contrast the harshness of mandatory LWOP for young offenders with the more moderate burdens of registration.
- Reaffirm Michigan’s commitment to individualized sentencing in extreme cases, even as it upholds mandatory SORA burdens here.
People v Dipiazza, 286 Mich App 137 (2009)
Referenced for its conclusion that registration can impair employment, housing, and “moving forward with life plans,” which informs the majority’s rehabilitation analysis.

B. Federal and Other-Jurisdiction Authorities

Smith v Doe, 538 US 84 (2003)
The U.S. Supreme Court held that Alaska’s pre-SORNA sex offender registration law was a civil, nonpunitive regulation, and therefore its retroactive application did not violate the Ex Post Facto Clause. Smith:
- Used the Mendoza‑Martinez factors.
- Emphasized limited reporting obligations (no in-person reporting), more modest public dissemination, and a closer fit between means and public-safety ends.
- Rejected analogies to colonial shaming and physical restraints.
The Kardasz majority distinguishes Smith on factual grounds and on the growth of internet-based publicity and empirical scholarship since 2003; Justice Zahra treats Smith as still controlling and analogous to Michigan’s regime.
Kansas v Hendricks, 521 US 346 (1997) & Kennedy v Mendoza‑Martinez, 372 US 144 (1963)
Provide the analytical and evidentiary framework for deciding whether a nominally civil scheme is actually punitive. Hendricks articulated the “clearest proof” standard. Mendoza‑Martinez articulated the multi-factor test for punitive effects.
Does #1‑5 v Snyder, 834 F3d 696 (6th Cir 2016); cert denied, 583 US 814 (2017)
The Sixth Circuit held that the 2006 and 2011 amendments to Michigan’s SORA were punitive and that their retroactive application violated the Ex Post Facto Clause. This federal ex post facto ruling was never reversed by the U.S. Supreme Court and heavily influenced Michigan’s own decision in Betts.
Does v Whitmer (Does III & IV), 751 F Supp 3d 761 (ED Mich 2024); 773 F Supp 3d 380 (ED Mich 2025)
A federal class action challenging the 2021 SORA on multiple grounds:
- Ex post facto: retroactive application of the 2021 SORA held punitive and unconstitutional under the federal Ex Post Facto Clause.
- Procedural due process: right to hearing for nonsexual convictions and out-of-state convictions.
- Equal protection: disparate treatment of out-of-state offenders.
- First Amendment: compelled speech (forced acknowledgments) and chilled speech (internet identifier requirements).
- Vagueness: several reporting provisions (phones, vehicles, travel plans, schooling) enjoined as unconstitutionally vague.
The Sixth Circuit has stayed those judgments pending appeal. The Michigan Supreme Court is not bound by these federal district court decisions but finds some of their reasoning and empirical evidence persuasive.
Other States’ Decisions
The opinion canvasses cases where state courts have found their registries punitive or ex post facto (e.g., Alaska, Maine, Indiana, Ohio, Oklahoma, Montana, New Hampshire). These illustrate a broader national trend of skepticism toward increasingly severe registries and are used by the majority to justify seeing SORA as punishment, though not to the point of striking it down as cruel or unusual.

V. The Majority’s Legal Reasoning

A. Step One: Is 2021 SORA “Punishment” or a Civil Regulation?

1. Legislative Intent

The Court starts with legislative intent, as required by Earl and Hendricks.

2021 SORA retains a statement of purpose: to assist law enforcement and the public in “preventing and protecting against the commission of future criminal sexual acts by convicted sex offenders” and to “monitor” such offenders. MCL 28.721a.
These are classic nonpunitive goals (public safety, monitoring), but:
- They also involve deterrence, which can be both punitive and regulatory.
- SORA is imposed as a consequence of conviction and enforced by criminal penalties for violations.

Following Lymon, the Court again holds that the Legislature intended SORA to be a civil, regulatory regime, especially given that some of the more obviously punitive features (like student safety zones) were removed in 2021.

2. Punitive Effects under the Mendoza‑Martinez Factors

Because legislative intent was civil, the question becomes whether SORA is nevertheless “so punitive in purpose or effect as to negate” that intent. The Court applies five of the Mendoza‑Martinez factors (as in Betts, Lymon, and Smith):

Historical analogues (has this type of measure been traditionally viewed as punishment?).
Affirmative restraint or disability.
Traditional aims of punishment (retribution, deterrence).
Rational connection to a nonpunitive purpose.
Excessiveness relative to that nonpunitive purpose.

a. Historical Tradition (Parole & Shaming)

The majority:

Agrees with Lymon that the 2021 SORA no longer resembles banishment, because student safety zones and geographic exclusion zones were removed.
Nonetheless finds clear resemblance to:
- Parole: frequent required contact with law enforcement, monitoring over long periods, and threat of imprisonment for noncompliance.
- Shaming: public web dissemination of detailed personal and offense information, mapped locations, and a “track offender” feature that allows the public to receive updates.

The Court underscores the difference between:

A one-off check of an official archive (Smith’s characterization of Alaska’s earlier registry), and
Michigan’s current, searchable, map-based, bulk-accessible, continuously updated website that can be used to stigmatize registrants in the “contemporary town square” of the internet.

Thus, this factor weighs in favor of finding SORA punitive.

b. Affirmative Disability or Restraint

The Court:

Rejects the idea that SORA’s effects are “minor and indirect.”
Emphasizes:
- Extensive personal information disclosure (home/work/school addresses, vehicles, physical identifiers, internet identifiers, etc.).
- Regular and sometimes quarterly in-person reporting (for Tier III), plus prompt in-person reporting upon changes in core life circumstances (residence, employment, schooling, name).
- Criminal penalties (including imprisonment) for “even technical violations” of reporting or fee requirements.

The majority distinguishes Smith v Doe on several factual grounds:

The Alaska scheme at issue in 2003 did not require in-person reporting.
The scale and social salience of the internet have dramatically increased since 2003, magnifying SORA’s real-world impact (e.g., vigilantism, harassment, housing/employment discrimination).
Research now documents that registry-related stigma and harassment are no longer rare anomalies but widespread collateral consequences.

The Court concludes that the combination of intrusive reporting and serious criminal penalties for noncompliance constitutes a significant affirmative restraint, again pointing toward punishment.

c. Traditional Aims of Punishment (Retribution & Deterrence)

The majority acknowledges that:

Deterrence is explicitly part of SORA’s stated purpose—“preventing” future criminal sexual acts.
Deterrence alone does not render a measure punitive (per Smith), but here the deterrent features are central rather than incidental.
The registry’s burdens are keyed to the fact of conviction, not to any individualized assessment of current risk or need, which resembles retribution for past acts more than risk-based regulation.

Given this structure, the Court finds that SORA promotes traditional penological goals—especially retribution—and that this factor also favors finding punishment.

d. Rational Connection to a Nonpunitive Purpose

Here, the majority agrees with Smith and prior Michigan cases:

There is unquestionably a rational connection between SORA and a legitimate, nonpunitive purpose:
- Assisting law enforcement in monitoring sex offenders.
- Providing the public with information that may help prevent future sexual victimization.
Courts do not require perfect tailoring; rationality is a “low bar.”

This factor cuts against finding SORA punitive and supports the Legislature’s civil characterization.

e. Excessiveness Relative to the Nonpunitive Purpose

This is where the majority’s analysis is most detailed and critical, and it draws heavily on empirical studies and risk-assessment tools.

(i) Offense-Based Tier System

SORA’s tiers (I, II, III) and corresponding durations (15, 25, life) are based solely on the offense of conviction and prior registrable offenses, mirroring federal SORNA.
The Court notes that:
- Michigan’s Department of Corrections uses scientifically validated instruments such as the Static‑99R to assess individualized recidivism risk for sex offenders.
- Empirical research indicates offense-based tiers correlate poorly with actual risk levels—some Tier I offenders have higher actuarial risk than some Tier III offenders, and overall the tiers both over- and under-include.
- Data from Michigan show an “inverse correlation” between tier and risk scores on Static‑99R (on average, Tier I registrants had higher risk scores than Tier III registrants).

The Court does not require the Legislature to adopt risk-based assessment systems, but it concludes that basing lifetime and long-term reporting burdens solely on offense categories, without accounting for time since release or demonstrated desistance, is excessive relative to the public-safety rationale.

(ii) Public Website and Overinclusion

The Court acknowledges rational reasons for a public website:
- Encouraging reporting by victims, including repeat victimization by known offenders.
- Providing affordable background-check capacity to the public.
However, studies from Oregon and Minnesota—where public listing is limited to high-risk offenders—suggest that:
- More targeted public notification schemes may reduce recidivism.
- Overly inclusive public registries may actually increase risk by destabilizing housing, employment, and support networks.
Michigan’s website publishes data on approximately 90% of non-incarcerated registrants, far greater than the 6–14% figures in Oregon and Minnesota.

The Court therefore characterizes the public web component as overinclusive and potentially counterproductive—another form of excessiveness.

(iii) Lengthy and Lifetime Reporting with Minimal Exit Options

Research on desistance suggests:
- High-risk offenders may need about 20 years offense-free in the community before their risk approaches that of the general population.
- Some low-risk offenders are effectively “desisted” at the time of release.
However:
- Many registrants (including elderly or disabled individuals with virtually no realistic risk of reoffending) remain on the registry for life.
- Tier III offenders—who include some repeat Tier II offenders and those with severe initial offenses—have very limited avenues for removal (e.g., certain juvenile adjudications and narrow “Romeo & Juliet” situations).
- There is no broad, individualized removal mechanism based on sustained offense-free living, age, infirmity, or professional risk assessments.

The Court sees the combination of long (often lifetime) reporting durations and lack of individualized exit as excessive in many cases, undermining the argument that the regime is purely regulatory.

Taking these aspects together, the Court concludes that the excessiveness factor supports a finding that the 2021 SORA is punitive, even though excessiveness is not extreme enough to make the entire regime unconstitutional as “cruel or unusual.”

f. Overall Punitive-Effects Conclusion

Weighing the factors:

Four factors (history/tradition, restraint, penological aims, excessiveness) point toward punishment.
One factor (rational connection to a nonpunitive purpose) supports the civil characterization.

On that basis, and under the “clearest proof” standard from Hendricks, the Court holds that:

The 2021 SORA is “so punitive either in purpose or effect as to negate the State’s intention to deem it civil.”

Thus, SORA is “punishment” for Article 1, § 16 purposes, even though it was framed as civil legislation.

B. Step Two: Is the 2021 SORA “Cruel or Unusual” Punishment?

Having labeled SORA “punishment,” the Court then asks whether the lifetime Tier III registration requirement is “cruel or unusual” either:

Facially—in all its applications, or
As applied to this defendant (a Tier III offender convicted of CSC-I against a child under 13).

The Court uses the Lorentzen/Bullock four-factor gross-disproportionality test.

1. Severity vs. Gravity of the Offense

Under Factor 1, the Court compares:

The harshness of the penalty: lifetime registration, quarterly in-person reporting if released, continuous update obligations, and associated collateral consequences.
The gravity of the offense: here, an aggravated sexual assault of a five-year-old (CSC-I under MCL 750.520b(1)(a) and (2)(b)), which the Court characterizes as “particularly grave.”

Key points:

The Court views SORA’s burdens as significant but moderate relative to the seriousness of Tier III offenses in general and this offense in particular.
The Court distinguishes this case from those involving mandatory life without parole for young adults (Parks, Taylor), where the harshness of LWOP was extreme relative to the penological justifications.
The Court notes that defendant provided no individualized evidence showing he is low-risk or desisted such that lifetime registration would be disproportionate as applied.

Factor 1 thus does not support a finding of gross disproportionality, either facially or as applied.

2. Comparison to Other Offenses in Michigan

Factor 2 requires an intrajurisdictional comparison—how does the punishment for this offense compare to:

Other SORA-eligible sexual offenses, and
Other serious non-sexual offenses that do not carry registration requirements.

The Court observes that:

The tier structure tracks offense severity in a rational, if imperfect, way:
- Tier III: most serious sex offenses (including CSC-I, assault with intent to commit penetration; repeat Tier II offenders) → lifetime registration / quarterly reporting.
- Tier II: e.g., allowing a child to engage in sexually abusive activity → 25 years / biannual reporting.
- Tier I: least serious registrable offenses (e.g., some instances of CSC-IV, possession of child sexually abusive material) → 15 years / annual reporting.
Compared to other grave non-sexual crimes (e.g., second-degree murder, terrorism, torture), which can yield life or lengthy prison terms, lifetime registration is serious but not obviously more severe than incarceration itself.

While the Court has already acknowledged that the tier system is “excessive” relative to nonpunitive goals when viewed from a regulatory design perspective, that excessiveness does not translate into constitutional disproportionality:

Tier III offenders have committed the most severe sex crimes or repeated serious sex crimes.
Their registration burden, relative to others in the same jurisdiction, is not grossly out of scale.

Therefore, Factor 2 does not favor a finding of gross disproportionality.

3. Comparison to Other States

Factor 3 examines whether Michigan’s scheme is outlier-level harsh compared to other states’ penalties for similar conduct.

The data, as the Court reads it, show:

Many states employ SORNA-aligned, offense-based, multi-tier systems similar to Michigan’s; Michigan is among a substantial plurality that:
- Impose lifetime registration for some categories of offenders.
- Do not generally provide a broad-based individualized removal mechanism for Tier III offenders.
Some states have moved toward:
- Risk-based classification.
- More robust removal opportunities.
- More limited public notification.

However, the record does not reveal a “clear national trend” away from regimes like Michigan’s—unlike, for example, the strong national movement away from juvenile LWOP in Parks and Stovall.

Accordingly, Factor 3 does not support a finding that Michigan is an outlier whose scheme is grossly excessive by national standards.

4. Rehabilitation and Penological Policy

Factor 4 is where SORA fares worst under Michigan’s constitutional tradition, which places distinct emphasis on rehabilitation:

SORA does not address underlying causes of offending (e.g., treatment, mental health, social supports).
SORA does not facilitate reintegration (e.g., by facilitating employment or housing).
On the contrary, the Court cites empirical work showing:
- Registration can harm financial stability, housing, and mental health.
- These harms can undermine law-abiding behavior and reintegration.
SORA fails to align burdens with individualized risk, which undercuts rehabilitative tailoring.

The Court thus concludes that Factor 4 weighs in favor of finding disproportionality.

5. Overall Proportionality Conclusion

Balancing all four factors, the Court holds:

Although the rehabilitative factor and some design excesses are troubling, the overall punishment imposed by Tier III registration is not grossly disproportionate:
- Not facially for all offenders in Tier III, and
- Not as applied to this defendant, whose offense is among the gravest.

Therefore:

The 2021 SORA does not violate Article 1, § 16, either facially or as applied.
Because Michigan’s constitutional standard is more protective, the Eighth Amendment claim also fails as a matter of course.

VI. The Separate Opinions in More Detail

A. Justice Zahra: SORA Is Not Punishment At All

Justice Zahra agrees entirely with the Court’s bottom-line disposition (SORA is not cruel or unusual; LEM issues not reached) but sharply disagrees with the majority’s premise that SORA is punishment.

Key themes:

Deference to Legislative Judgment
He stresses that it is an “extraordinarily high” bar to overturn a legislature’s civil label and that the “clearest proof” standard is not met. The majority’s repeated second-guessing of legislative policy choices (on risk, tiers, and web publication) is, in his view, an improper judicial intrusion.
Reliance on Smith v Doe
Zahra leans heavily on Smith:
- He argues that, like Alaska’s scheme, Michigan’s registry is fundamentally designed to disseminate accurate conviction information for public safety—not to shame or physically restrain.
- He rejects the majority’s analogy to colonial shaming and sees modern internet-based publication as constitutionally indistinguishable from other forms of disseminating public records.
Affirmative Restraint
He views the reporting obligations and web listing as far from the “paradigmatic” restraint of imprisonment and notes that registrants are still “free to live where they wish and to live and work as other citizens.” He distinguishes parole/probation (with intensive supervision, conditions, and revocation) from SORA’s more limited reporting.
Traditional Aims of Punishment
He acknowledges some deterrent effect but insists that deterrence alone does not make a scheme punitive. He denies that SORA meaningfully promotes retribution, viewing it instead as a rational classification and disclosure system.
Excessiveness
In his view, SORA is a reasonable means of achieving its nonpunitive purpose:
- Tiered reporting corresponds to seriousness of the underlying offense.
- Lifetime registration for Tier III offenders is appropriate because they are the “most dangerous subset of sex offenders.”
- Public web listing is not overinclusive but rationally targeted at those convicted of sexual crimes.
Warning about Systemic Consequences
Zahra suggests that declaring SORA punitive (first in Lymon for nonsexual offenses and now in Kardasz for sexual offenses) threatens the legal foundation of public registries generally and “erodes the safety and well-being of Michigan families” by judicially weakening preventive tools.

B. Justice Bernstein: Avoid the Punishment Question

Justice Bernstein concurs in the judgment and in almost all of the majority’s reasoning but would:

Avoid deciding whether SORA is “punishment.”
Simply assume for purposes of analysis that, even if it is punishment, it is not “cruel or unusual” in this case.

His short separate opinion reflects a jurisprudential preference for minimum necessary constitutional decision-making (i.e., not reaching issues unnecessary to the outcome).

VII. Clarifying Key Legal Concepts

A. “Punishment” vs. Civil Regulation

Not every burdensome law is a “punishment” in the constitutional sense. Courts distinguish:

Criminal punishment: jail, fines, probation, etc., imposed as retribution or condemnation.
Civil regulation: licensing, registration, professional discipline, etc., imposed to manage risk, protect the public, or regulate conduct.

The line matters because:

Ex post facto prohibitions apply only to retroactive punishments.
Cruel or unusual / cruel and unusual protections primarily apply to punishments.

The Earl/Hendricks/Mendoza‑Martinez framework is the tool used to decide whether a scheme that calls itself “civil” is actually punitive in fact.

B. Michigan’s “Cruel or Unusual” vs. Federal “Cruel and Unusual”

The Michigan Constitution forbids “cruel or unusual punishment.”
The Eighth Amendment forbids “cruel and unusual punishments.”

Michigan courts have long held that this textual difference gives broader protection in some contexts than the federal Eighth Amendment, particularly in proportionality and juvenile sentencing cases. Thus, if a punishment survives under Michigan’s more protective standard, it almost always survives federal scrutiny.

C. Facial vs. As-Applied Challenges

Facial challenge: asserts that a law is unconstitutional in all or virtually all of its applications—there is no set of circumstances in which it can be validly applied.
As-applied challenge: concedes that the law may be valid in general but claims the law is unconstitutional as enforced against a particular person or in specific circumstances.

In Kardasz:

The Court rejects the facial challenge because some Tier III offenders’ crimes (including this one) are so serious that lifetime registration is not grossly disproportionate.
The Court rejects the as-applied challenge because defendant did not produce individualized risk evidence and his crime is at the extreme end of seriousness.

D. Mendoza‑Martinez Factors (Punitive Effects Test)

These are the key guideposts for determining whether a scheme labeled “civil” has punitive effects:

Has it traditionally been regarded as a punishment (e.g., banishment, shaming, probation/parole analogues)?
Does it impose significant restraints or disabilities beyond mere inconvenience?
Does it serve goals such as retribution or deterrence (classic aims of punishment)?
Is it rationally connected to a legitimate nonpunitive purpose?
Is it excessive in relation to that purpose?

No single factor is dispositive; the analysis is holistic. The “clearest proof” standard sets the overall burden of persuasion on the challenger.

E. Lorentzen/Bullock Gross-Disproportionality Test

To decide whether a punishment is “cruel or unusual,” Michigan courts look at:

Harshness vs. gravity of the offense.
Comparisons to other offenses within Michigan.
Comparisons to other states for the same offense.
Rehabilitation and broader penological policy.

The test does not require perfect proportionality, only that the penalty not be so extreme as to be “grossly disproportionate” to the offense.

F. SORA Tiers in Brief

Tier I (least restrictive):
- 15 years of registration.
- Annual in-person verification.
- Can typically petition for removal after 10 years.
Tier II:
- 25 years of registration.
- Biannual in-person verification.
Tier III (most restrictive):
- Lifetime registration.
- Quarterly in-person verification.
- Very limited removal possibilities (primarily certain juvenile adjudications and specific “Romeo & Juliet” situations).
- Includes CSC-I and other serious penetrative or violent sex offenses, as well as some repeat offenders.

VIII. Likely Impact and Future Directions

A. Consequences of Labeling SORA as “Punishment”

The Court’s explicit holding that the 2021 SORA is punishment has several important downstream effects:

Ex Post Facto Analysis
Because SORA is punishment, any retroactive application of its burdens must satisfy state and federal ex post facto prohibitions. That question was not raised in Kardasz (and had already been partially addressed in Betts), but it remains critical in ongoing litigation such as Does v Whitmer.
Sentencing and Plea Bargaining
Courts and counsel may need to treat registration more overtly as part of the “punishment package,” not a mere civil collateral consequence. This could:
- Influence plea advice requirements and voluntariness of guilty pleas.
- Influence proportionality review if prison terms plus SORA burdens together are argued to be excessive in a given case.
Scope of Michigan’s Cruel-or-Unusual Review
The decision reaffirms that collateral regulatory measures like registries, when sufficiently burdensome, can fall under Article 1, § 16—even if they do not involve incarceration in the traditional sense.

B. Space Left for Future Challenges

Although the Court upholds SORA, it explicitly identifies several structural features as excessive or problematic:

Offense-based tiers that correlate poorly with actual risk.
Broad, nearly universal public web listing, rather than risk-based public notification.
Lifetime registration with no general removal mechanism for high-risk categories, even after decades of offense-free living.

These observations, though not converting into a finding of “cruel or unusual” punishment here, effectively invite:

Legislative reform:
- Incorporating validated risk assessments (e.g., Static‑99R) into tier assignment and removal decisions.
- Narrowing public web publication to high-risk offenders and/or time-limited listings.
- Establishing meaningful petition procedures for removal or downgrading after long periods of offense-free behavior and demonstrated desistance.
Targeted constitutional challenges:
- Future as-applied challenges where defendants present strong individualized evidence of low risk after long offense-free periods.
- Continued litigation over due process (notice and hearing), equal protection, First Amendment, and vagueness, as in Does v Whitmer.

C. LEM and Fourth Amendment Issues Remain Open

By denying leave on LEM issues, the Court:

Leaves intact lower court precedent on LEM (e.g., Hallak), at least for now.
Defers for another day:
- Whether lifetime GPS monitoring of sex offenders constitutes a continuous “search” requiring reasonableness analysis (as held by some other states such as North Carolina and Georgia).
- Whether LEM is itself cruel or unusual punishment under the Michigan Constitution.

Given the Court’s willingness to recognize SORA as punishment, challenges to LEM as part of the punishment package—especially for long terms or life—will likely return in future cases, potentially with more developed records on privacy, proportionality, and rehabilitation.

D. Relationship to Federal Law and National Trends

Kardasz continues a broader trend among state supreme courts and lower federal courts toward a more skeptical view of modern, expansive registries than that expressed in the U.S. Supreme Court’s 2003 Smith v Doe decision:

Michigan now joins jurisdictions that have explicitly deemed their registries punitive (at least in some applications), even while upholding them in part.
The decision may influence other states, especially those with SORNA-based schemes, to reconsider whether their registries should be treated as civil or criminal for state constitutional purposes.

However, by upholding SORA against cruel-or-unusual challenges, Michigan remains aligned with many states that accept rigorous registration regimes as constitutionally permissible, particularly for the most serious sexual offenses.

IX. Conclusion

People v. Kardasz is a pivotal decision that both expands and constrains the constitutional status of Michigan’s sex offender registration system.

On one hand, the Court decisively holds that the 2021 SORA is punishment under the Michigan Constitution, rejecting the notion that it is a mere civil regulatory tool. This has important implications for ex post facto analysis, proportionality review, and how courts and practitioners conceptualize registration in sentencing and plea contexts.
On the other hand, the Court holds that SORA—at least as applied to Tier III sex offenders like Kardasz—is not cruel or unusual punishment, either facially or as applied. The burdens of lifetime registration and reporting, though heavy and imperfectly tailored, are not “grossly disproportionate” to the gravity of the underlying crimes in the Court’s view.

The separate opinions sharpen the stakes: Justice Zahra warns against judicial encroachment on legislative prerogatives and insists SORA remains a civil regime; Justice Bernstein prefers to avoid deciding more than necessary. The majority nonetheless stakes out a middle path—recognizing the punitive reality of modern registries while stopping short of striking them down under proportionality principles.

Going forward, Kardasz will serve as a foundational precedent:

For legislators considering reforms to make SORA more risk-based, less overinclusive, and more consistent with rehabilitative goals.
For courts evaluating ex post facto, due process, and tailored as-applied challenges to particular SORA provisions or to long-term LEM.
For scholars and practitioners grappling with the line between civil regulation and punishment in an era of expansive collateral consequences.

In sum, the Michigan Supreme Court has acknowledged that sex offender registration is, in substance, part of the punishment imposed on offenders. But it has also signaled that the Constitution does not yet prohibit the particular form of punishment embodied in the 2021 SORA, at least for the most serious sexual crimes. The real work of refining—or reimagining—this regime now falls squarely back on the Legislature and on future litigants armed with more specific, individualized records.