Response to the Supreme Court's Order of January 7, 2009
PER CURIAM.In a petition for hearing currently pending before the Alaska Supreme Court, Byron Charles challenges his conviction for failing to register as a sex offender as required by Alaska's sex offender registration act, AS 12.68.
(This Court affirmed Charles's conviction on direct appeal. See Charles v. State, Alaska App. Memorandum Opinion No. 5277 (Nov. 28, 2007); 2007 WL 4227885.)
After Charles filed his petition for hearing, the supreme court issued its decision in Doe v. State, 189 P.3d 999 (Alaska 2008). In Dog, the supreme court held that the requirements of the sex offender registration act constitute "punishment" for purposes of the ex post facto clause of the Alaska Constitution (Article I, Section 15), and therefore the sex offender registration act can not lawfully be applied to defendants whose offenses predate the effective date of the act (August 10, 1994). Id. at 1018-19.
Charles committed his underlying sex offense in the 1980s, before the enactment of the sex offender registration act1 The supreme court now confronts the question of whether Charles is entitled to claim the benefit of the decision in Doe.
In an order dated January 7, 2009, the supreme court directed us to consider certain questions relating to whether Charles can claim the benefit of Doe.
First, the supreme court asks us to address the issue of whether Charles has waived any ex post facto challenge to the sex offender registration act-and, if so, whether (as a legal matter) ex post facto challenges to a statute are ever waivable.
Second, the supreme court asks us to give our opinion as to whether Alaska should modify the current rule governing the retro-activity of court decisions-the rule announced in Judd v. State, 482 P.2d 273 (Alaska 1971)-by adopting the federal principle of retroactivity announced by the United States Supreme Court in Griffith v. Kentucky, 479 U.S. 314, 107 S.Ct. 708, 93 L.Ed.2d 649 (1987) the principle that new constitutional rulings should always be applied to all defendants whose convictions are not yet final when the ruling is announced.2
*781Finally, the supreme court asks for our opinion as to whether Charles's conviction for failure to register as a sex offender should be set aside under the Griffith principle of ret-roactivity.
Because the supreme court has directed us to consider these questions but has not entered a final order with respect to Charles's petition for hearing, we do not interpret the supreme court's order as giving this Court jurisdiction to independently decide these issues. Rather, we believe that the supreme court has merely interrupted its own consideration of Charles's case to seek our input and recommendations on these issues.
Given the importance of these legal issues, not only for Charles's case but for future cases as well, we asked the parties to file supplemental briefs on the questions posed by the supreme court, and we thank the parties for their thoughtful input.
Because the supreme court has not asked us to decide these issues, but rather to offer our analysis and advice, we have not written this response the way we would normally craft a judicial opinion. Rather, our response is more in the nature of a memorandum: we discuss the existing law, we discuss how that law might apply to the issues potentially raised in Charles's case, and we discuss why the supreme court might not need to resolve all of these legal issues in Charles's case.
Has Charles waived or forfeited any ex post facto claim?
The question of whether a person has "waived" a constitutional right is somewhat ambiguous, because courts speak of waiver in two different contexts. On the one hand, a person can "waive" a right by knowingly choosing to relinquish that right, or by knowingly choosing not to exercise that right. On the other hand, courts often use the term "waiver" when they refer to a person's forfet-ture of a right by failing to exercise or claim the benefit of that right-even when the person has made no conscious decision to forego the right, and has simply neglected to assert it.
As the State concedes, there is nothing in the record of this case to suggest that Charles has ever knowingly chosen to forego his rights under the ex post facto clause. Thus, there is no need to address the question of whether ex post fucto rights are "waivable" in this sense-ie, no need to decide whether the law might allow a defendant to knowingly choose to forego the protections of the ex post facto clause.
We note, however, that there are potentially instances where it might be to a defendant's advantage to waive ex post facto rights. For instance, in the present case, Charles was charged with "failure to register" as a sex offender even though, as a factual matter, he did register. The problem was that he supplied a false address when he registered. See Charles v. State, 2007 WL 42278835 at *1.
The State charged Charles with a class A misdemeanor under AS 11.56.840(a)(2). But under these circumstances, the State might conceivably have charged Charles with a more serious crime. Under AS 12.63.010(e), the registration forms required by Alaska's sex offender registration act "must be sworn to by the offender ... and [must] contain an admonition that a false statement shall subject the offender ... to prosecution for perJury."
Perjury is a class B3 We can imagine that a defendant in Charles's position might wish to forego their ex post facto defense to the misdemeanor charge of "failure to register" in exchange for the government's promise not to pursue a perjury prosecution. - For this reason, we believe it would be inadvisable to rule that a defendant can never voluntarily relinquish the protections afforded by the ex post facto clause.
(We note that this Court has already ruled that a defendant may knowingly relinquish *782the protections of the double jeopardy clause.4
This leaves the issue of whether Charles may have forfeited his potential ex post facto defense by failing to assert it until now.
Several federal and state courts have ruled that ex post facto protections can be forfeited by failing to assert them.5 Texas, however, has ruled that ex post facto protections can be neither waived nor forfeited, because the ex post fucto clause is "a categorical prohibition directed by the people against their government".6 The Texas Court of Criminal Appeals reasoned that defendants should not be permitted to waive the protections of the ex post facto clause "any more than they may consent to be imprisoned for conduct which is not a crime.7
However, even in jurisdictions where ex post facto protections are forfeited by inaction, the fact that a defendant has forfeited an ex post facto claim by failing to raise the claim in the trial court does not mean that the defendant is completely barred from seeking relief. Appellate courts that apply a rule of forfeiture to an unpreserved ex post facto claim generally still allow a defendant to litigate the claim on appeal under the rubric of "plain error" or "manifest injustice".8
We ourselves have repeatedly applied the plain error rule to claims of constitutional error that were not properly preserved in the trial court proceedings-most recently, in a series of sentencing cases arising from the United States Supreme Court's decision in Blakely v. Washington, 542 U.S. 296, 124 S.Ct. 2531, 159 L.Ed.2d 403 (2004).
In Blakely, the Supreme Court held that defendants are entitled to a jury trial, and entitled to demand proof beyond a reasonable doubt, on any issue of fact which, if found in the government's favor, would increase the maximum sentence for their crime. Because Blakely announced a new rule of federal constitutional law, Griffith v. Kentucky governed the retroactive application of Blakely. In other words, any defendant whose conviction was not yet final (.e, was still on direct review) at the time Blakely was decided could retroactively claim the benefit of Blakely.
But even though any defendant whose appeal was still pending could retroactively claim the benefit of Blakely, this Court repeatedly held that defendants who failed to preserve a Blakely claim in the trial court had to demonstrate not only that Blakely was violated in their case, but also that the Blakely violation amounted to plain error.9
This use of the phrase "plain error" actually refers to a concept that is somewhat different from what courts normally mean when they say "plain error". A claim of plain error typically includes an assertion that the lower court overlooked, or failed to remedy, an error that would have been obvious to any competent judge.10 But when a *783claim of error rests on the retroactive application of a new constitutional rule, no one can say that the trial court judge failed to perceive an obvious error-because, at the time of the lower court proceedings, the law was different.
In such cases, the notion of "plain error" is retrospective. The question is whether, in light of the new constitutional rule, we can mow see that there was an obvious flaw in the lower court proceedings, and that fallure to correct this flaw would perpetuate manifest injustice.
Thus, for instance, this Court repeatedly held that even though a defendant was wrongfully deprived of a jury trial on an aggravating factor that increased the permissible maximum sentence for the defendant's crime, this Blakely violation did not constitute plain error when the evidence concerning the existence of the aggravator was not subject to reasonable dispute-in other words, when there was no reasonable possibility that a jury would have found in the defendant's favor even if the issue had been submitted to a jury.11
In the present case, even though the State argues that Charles has forfeited his ex post facto claim by failing to raise an ex post facto argument in the trial court, the State acknowledges that Charles may still pursue his ex post fucto claim under the rubric of plain error-and that Charles would be entitled to relief if he can show that there was a "clear" violation "of [his] substantial rights" that prejudiced "the [fundamental] justice or integrity of the proceedings".
It is undisputed that Charles committed his underlying sex offense before the legislature enacted Alaska's sex offender registration law. Thus, even though Charles failed to raise an ex post fucto objection in the trial court, he would nevertheless be entitled to relief from his failure-to-register conviction if (1) the supreme court's decision in Doe applies retroactively to Charles, and if (2) the prosecution of Charles for failure to register as a sex offender, in violation of the ex post facto clause, amounts to "plain error"-that is, if the ex post facto violation was "so prejudicial to the fairness of the proceedings that ... failure to correct it would perpetuate manifest injustice.12
The State's arguments that there is no plain error in Charles's case because (1) Doe was wrongly decided, or (2) Doe is not binding precedent
Before proceeding further, we must address two arguments that the State raises. Both of these arguments are aimed at showing that, regardless of the law pertaining to retroactivity, Charles is not entitled to relief.
First, the State argues that Doe was wrongly decided.
Second, the State argues that, even if Doe was correctly decided, it is not "precedent"that is, it does not establish a rule of decision for other cases-because only three members of the supreme court participated in Doe (a bare quorum under Appellate Rule 105(a)), and because Doe was decided by a two-to-one vote.
The State relies on Appellate Rule 106(b), which was enacted by the supreme court in the summer of 2011, three years after the court issued its decision in Doe.13 Under Rule 106(b), a two-to-one decision of the supreme court applies only to the resolution of that particular appeal, "and [does] not have prece-dential effect".
The State argues that Appellate Rule 106(b) has retroactive effect-that the enactment of this rule stripped Doe of its status as binding precedent. Accordingly, the State argues that, regardless of the law governing the retroactivity of court decisions, neither Charles nor any other defendant is entitled to relief based on Doe.
For two reasons, we decline to address the State's arguments.
*784First, these issues fall outside the seope of the questions presented to us in the supreme court's order of January 7, 2009. Indeed, with regard to the State's argument based on Appellate Rule 106(b), we note that the supreme court's directive to this Court was issued two and a half years before the supreme court enacted Appellate Rule 106(b).
Second, because Charles's case is still pending before the supreme court, and because the supreme court is about to resume active consideration of Charles's case, it seems to us that there would be little value in having this Court offer an opinion on the underlying merits of the decision in Dog, or in having this Court try to assess how the supreme court would interpret Appellate Rule 106(b) as it relates to Doe.
(Compare AS 22.05.015(b), which authorizes this Court to ask the supreme court to assume jurisdiction of a case that would normally be within this Court's jurisdiction if "the case involves ... an issue of substantial public interest that should be determined by the supreme court.")
We now return to the issues raised in the supreme court's order.
Should Alaska adopt the retroactivity rule set forth in Griffith v. Kentucky?
The supreme court has asked for our opinion as to whether Alaska should modify our current rule governing the retroactivity of court decisions-the rule announced in Judd v. State, 482 P.2d 273 (Alaska 1971)-by adopting the federal principle of retroactivity announced by the United States Supreme Court in Griffith v. Kentucky, 479 U.S. 314, 107 S.Ct. 708, 93 L.Ed.2d 649 (1987).
Under the current Judd rule, the issue of retroactivity is assessed by weighing three factors: (1) the purpose to be served by the new rule; (2) the extent to which law enforcement authorities have relied on the old rule; and (3) the degree to which retroactive application of the new rule would disrupt the administration of justice.14 Under Griffith, on the other hand, any new constitutional rule must be applied to all defendants whose convictions are still on direct review at the time the rule was announced.15
Charles's supplemental brief to this Court contains a lengthy discussion of the problems and potential injustices created by the Judd rule, and the benefits of the Griffith rule. The State agrees that "there may be reasons for rejecting the ... Judd standard."
However, as we explain in the next section of this response, we conclude that our supreme court's decision in Doe must be applied retroactively to Charles under either Alaska's current Judd rule of retroactivity or the Griffith rule of retroactivity. We therefore conclude that there is no immediate need for the supreme court to decide whether to adopt the Griffith principle of retroac-tivity.
Although all members of this Court agree that there is no need to decide this question of retroactivity law, Judge Coats has written a concurring opinion in which he discusses the perceived advantages of the Judd rule over the Griffith rule. For this reason, we (the other two members of this Court) believe that it is worthwhile to mention the policies that led the United States Supreme Court to adopt the Griffith rule.
Initially, the federal test for retroactivity was the same as our Judd test-because the Judd rule of retroactivity was taken directly from the rule formulated by the United States Supreme Court in Linkletter v. Walker, 381 U.S. 618, 636, 85 S.Ct. 1731, 1741, 14 L.Ed.2d 601 (1965), and reiterated two years later in Stovall v. Denno, 388 U.S. 293, 297, 87 S.Ct. 1967, 1970, 18 L.Ed.2d 1199 (1967). As we have explained, under this test, the retroactivity of a court decision is assessed by weighing the purpose of the new rule, the degree to which people reasonably relied on the old rule, and the degree to which retroac*785tive application of the new rule would disrupt the administration of justice.
Originally, the Linkletter/Stovall rule of retroactivity applied equally to defendants whose convictions were still being challenged on direct review when the new rule was announced, and to defendants whose convictions had already become final.16 But in United States v. Johnson, 457 U.S. 537, 102 S.Ct. 2579, 73 L.Ed.2d 202 (1982), the Supreme Court reconsidered this unified approach to retroactivity and declared that different policies should govern the assessment of whether a new rule should apply retroactively to defendants whose convictions were not yet final when the decision was announced. Id., 457 U.S. at 548, 102 S.Ct. at 2586.
In particular, the Supreme Court adopted the view that, after a new rule has been announced in one defendant's case, that same rule must be applied to any "similarly situated defendant" whose conviction is not yet final unless there is "a principled reason for acting differently." Id., 457 U.S. at 561-62, 102 S.Ct. at 2593.17
In its Johnson decision, the Supreme Court allowed one major exception to this principle that new rules should be applied retroactively to all defendants whose cases were currently pending on direct review at the time the rule was announced: the Supreme Court declared that a new rule would not be applied retroactively if the new rule constituted a "clear break" from the Court's past precedents. Id., 457 U.S. at 549-550, 102 S.Ct. at 2586-2587.
But five years later, in Griffith v. Kentucky, 479 U.S. 314, 107 S.Ct. 708, 93 L.Ed.2d 649 (1987), the Supreme Court abandoned this "clear break" exception and declared that all new rules would apply retroactively to defendants whose convictions were not yet final when the new rule was announced.
In Griffith, the Supreme Court concluded that it was fundamentally unfair to "fish[ ] one case from the stream of appellate review" and employ that case "as a vehicle for announcing new constitutional standards", but then refuse to apply the new standards to all the other defendants who were waiting for their appeals to be heard. Id., 479 U.S. at 828, 107 S.Ct. at 713. The Court concluded that a refusal to apply the new rule to all defendants whose convictions were not yet final would create an unacceptable inequity-because one defendant (the defendant whose case was chosen by the appellate court as a vehicle for announcing the new rule) would be the "chance beneficiary" of the new rule, while all other similarly situated defendants would see their same claims denied. 479 U.S. at 828 & 327, 107 S.Ct. at 713 & 715.
The Supreme Court acknowledged that, when a new rule constitutes a "clear break" from earlier precedent, the second and third Linkletter factors-reasonable reliance on the old rule, and the probable adverse effect on the administration of justice-were likely to weigh heavily against retroactive application. Griffith 479 U.S. at 326-27, 107 S.Ct. at 715. Nevertheless, the Court concluded that, even in such instances, it was unacceptably unfair to fail to apply a new rule retroactively to all defendants whose cases were still on direct appeal. Id., 479 U.S. at 827-28, 107 S.Ct. at 715-16.
We acknowledge that almost any rule of limited retroactivity carries the seeds of arbitrariness and unfairness. For instance, Griffith declares that there is one rule of ret-roactivity for defendants who are still seeking appellate review of their convictions when a new rule is announced, and another rule of retroactivity for defendants whose convictions are already final when the new rule is announced. This means that an appellate court is required to apply the new rule retroactively to all defendants who are still waiting for the decision of their petition for hearing (in our state court system) or their petition for certiorari (in the federal system), but to deny retroactive application to defendants whose petitions for discretion*786ary review were denied just a few days earlier.
Moreover, in the case of new rules that are a "clear break" from earlier precedent, the Griffith rule means that some defendants will be denied relief because (1) they raised the same claim on direct appeal but lost, and then (2) they failed to pursue a petition for discretionary review after their attorney advised them-competently-that they had little or no chance of prevailing on such a claim.
As we explained earlier, we conclude that Charles's case does not require our supreme court to decide whether to adopt the Griffith rule of retroactivity-because Charles is entitled to retroactive application of the Doe decision regardless of whether our supreme court applies the Griffith rule or the current Judd rule. We have presented this extensive examination of Griffith, not in an effort to convince our supreme court to adopt the Griffith rule of retroactivity, but rather to elucidate the scope of the Griffith rule and the policies behind it, so as to round out the thoughtful views expressed by Judge Coats in his concurrence.
We additionally note-for purposes of clarification-that even if Alaska adopted the Griffith rule, this would not necessarily entail a complete abrogation of Alaska's current Judd rule of retroactivity.
Under Griffith, a new rule applies retroactively to defendants whose convictions are not yet final at the time the new rule is announced. But Griffith does not answer the question of whether a new rule should be applied retroactively to defendants whose convictions are already final when the new rule is announced.
Under federal law, this latter situation is governed by the rule of retroactivity announced in Teague v. Lane, 489 U.S. 288, 109 S.Ct. 1060, 103 L.Ed.2d 334 (1989). Under Teague, a new rule will be applied completely retroactively (that is, even to defendants whose convictions are already final) only if the new rule is a rule of substantive law (i.e., a rule that limits the authority of the government to prosecute or punish certain conduct), or if the new rule is a "watershed" procedural rule that implicates the fundamental fairness of the criminal proceeding or the fundamental accuracy of the fact-finding process.
The seope of the Teague retroactivity rule is explained in Smart v. State (Smart I ), 146 P.3d 15 (Alaska App.2006):
Under ... Teague, a [Judicial] decision creates a new rule "when it breaks new ground or imposes a new obligation on the States or the Federal Government." [Teague, 489 U.S. at 301, 109 S.Ct. at 1070.] When the [judicial] opinion creates a new rule, the rule applies to previously final judgments only in limited cireum-stances. For example, new substantive rules which decriminalize a class of conduct or prohibit capital punishment for a class of defendants generally apply [completely] retroactively. These new substantive rules apply retroactively because there is a risk that the defendant was convicted for an act that is not criminal or [that the defendant] faces a punishment that is not allowed by law. However, new procedural rules generally do not apply retroactively. New procedural rules merely raise the possibility that someone convicted with the use of the invalidated procedure might have been acquitted otherwise. Because of this more speculative connection to innocence, [the courts] give retroactive effect to only a small set of watershed rules of criminal procedure implicating the fundamental fairness and accuracy of the criminal proceeding.
Smart I, 146 P.8d at 40 (concurring opinion of Judge Stewart) (omitting several footnotes that cite, or quote from, the United States Supreme Court's decision in Schriro v. Summerlin, 542 U.S. 348, 351-52, 124 S.Ct. 2519, 2522-23, 159 L.Ed.2d 442 (2004)).
But the states are not required to follow the Teague rule of retroactivity, so long as their rules of retroactivity are at least as protective of defendants' federal rights as the Teague rule.18 And Alaska law is currently unsettled as to what rule of retroactivity governs situations where a defendant's conviction is already final when a new rule is announced.
*787The Judd rule is the Alaska Supreme Court's last pronouncement on this question, and this is the rule that the supreme court applied in State v. Smart (Smart II), 202 P.3d 1130, 1138-39 (Alaska 2009), its most recent decision addressing the question of retroactivity.
However, AS 12.72.010(7)-a portion of the statute governing the seope of post-convietion relief in Alaska-contains a different set of rules that govern the retroactivity of judicial decisions when these decisions are invoked for the purpose of collaterally attacking a criminal conviction that is already final. In prior appellate litigation, the State has taken the position that AS 12.72.010(7) was intended to codify the Teague rule of retroac-tivity, and thus the Teague test is the governing rule of retroactivity for post-convietion relief litigation in Alaska.19
As this Court noted in Smart I, 146 P.3d at 29, the conflict between the Judd rule and the provisions of AS 12.72.010(7) potentially raises a question under the doctrine of separation of powers: specifically, whether the legislature's enactment of rules governing the retroactivity of judicial decisions improperly encroaches on a matter entrusted to the judicial branch. However, there is no need to resolve this question, because it is moot under the facts of Charles's case-since Charles's conviction is not yet final.
We do note, however, that the Alaska Supreme Court has declared that the Judd rule of retroactivity is at least as protective of defendant's rights as the Teague rule. This issue came up in Smart IL.
The primary question presented in Smart II was whether the right to jury trial announced in Blakely v. Washington should be applied retroactively to defendants whose convictions were already final when Blakely was decided. Our supreme court recognized that, because Blakely announced a new rule of federal law (an expansion of the scope of the Sixth Amendment right to jury trial), state courts were obliged to make Blakely retroactive at least to the extent guaranteed by the Teague test.20
But rather than adopting the Teague test as the standard for retroactivity under Alaska law, the supreme court adhered to the Judd test. The supreme court was able to do this because the court declared that the Judd rule of retroactivity was at least as protective of defendants' rights as the Teag-ue test. Smart II, 202 P.3d at 1188-39.
Why we conclude that Charles is entitled to retroactive application of the Doe decision under the Judd rule of retroactivity as well as the Griffith rule of retroactivity
Under the Griffith rule of retroactivity, constitutional decisions must be applied retroactively to all defendants whose convie-tions are not yet final at the time the decision is announced. Although this Court had already affirmed Charles's conviction on direct appeal when the Alaska Supreme Court announced its decision in Doe, Charles was still in the process of petitioning the supreme court to review this Court's affirmance of his conviction. Thus, Charles's conviction was not yet final when the supreme court issued its decision in Doe.21 Under the Griffith rule, it is clear that Charles would be entitled to retroactive application of the Doe decision.
Although the question of retroactivity is more complicated under the Judd test, the result is the same. As we are about to explain, the Doe decision would be given full retroactivity under the Teague test (if the Doe decision rested on federal law). And because, in Smart ILI, our supreme court declared that the Judd test was at least as protective of defendants' rights as the Teag-ue test, it follows that Doe should receive full retroactivity under Judd.
To back up a bit: The Judd test for deciding whether a new rule should be applied retroactively involves the weighing of three *788factors: (1) the purpose to be served by the new rule; (2) the extent to which law enforcement authorities have relied on the old rule; and (8) the disruptive effect that retroactive application of the new rule would have on the administration of justice.22
However, our supreme court has declared that the first Judd factor-the purpose to be served by the new rule-will take precedence over the other two factors, and will require retroactive application of a new constitutional rule, if the primary purpose of the new rule is to enhance the truth-finding function of criminal trials, thus helping to ensure fair and accurate verdicts. Rutherford v. State, 486 P.2d 946, 952-53 (Alaska 1971).
The new rule at issue in the present case-the Doe court's pronouncement that sex offender registration is a "punishment" for purposes of Alaska's ex post fucto clause-is not a rule that enhances the truth-finding function of criminal trials. However, the Doe rule is likewise aimed at protecting defendants from unjust conviction and punishment.
Under our tripartite system of government, it is the legislative branch that defines crimes and establishes the punishments for those crimes.23 The Alaska Supreme Court has explicitly recognized this principle:
Save only as limited by constitutional safeguards, the legislature may choose any reasonable means to protect the people from the violation of criminal laws. In general, the comparative gravity of offenses and their classification and resultant punishment is for legislative determination.
Alex v. State, 484 P.2d 677, 685 (Alaska 1971) (citation omitted).
But the ex post facto clause is one of the constitutional safeguards that limits the legislature's authority in this area. With respect to criminal punishments, the ex post facto clause strictly forbids the legislature from "impos[ing a] punishment more severe than the punishment [that was] assigned by law [to the offense] when the [defendant committed the] act to be punished". Weaver v. Graham, 450 U.S. 24, 30, 101 S.Ct. 960, 965, 67 LEd.2d 17 (1981), quoted with approval in State v. Creekpaum, 753 P.2d 1139, 1141 (Alaska 1988).
As explained in Weaver and in Creekpaum, the constitutional prohibition on ex post facto punishments is not premised on an individual defendant's right to a lesser punishment. Rather, the ex post facto clause is a restraint on the authority of the legislature itself: the legislature is forbidden from "increas[ing] punishment beyond what was prescribed when the crime was consummated." Weaver, 450 U.S. at 30, 101 S.Ct. at 965; Creekpaum, 753 P.2d at 1141.
Because the Alaska Supreme Court, in Doe, declared that sex offender registration is a "punishment" for ex post facto purposes, it necessarily follows that it is unjust-in a fundamental, constitutional sense-to impose that punishment on any defendant whose offense predates the enactment of the sex offender registration law. Accordingly, this appears to be the type of situation where the first Judd factor takes precedence over other considerations and demands retroactive application of Doe.
Moreover, even if we applied the pre-sumedly more restrictive federal test for ret-roactivity announced in Teague v. Lane,24 our supreme court's decision in Doe would be given complete retroactive effect.
The Teague rule of retroactivity is premised on the principle that if a criminal convietion has become final, and if the litigation of that criminal case took place in compliance with the law that was in existence at the time, then there should be very few circumstances which compel a re-examination of the criminal judgement based on new developments in constitutional law.25 Under Teague, *789a retroactive attack on a final criminal judgement is allowed only if (1) the new constitutional rule is a substantive rule that "[puts] certain kinds of primary, private individual conduct beyond the power of [the states to regulate through eriminal legislation]", or if (2) the new constitutional rule is a "watershed" procedural rule that "requires the observance of ... procedures that are implicit in the concept of ordered liberty26
But even under the restrictions of the Teague test, our supreme court's decision in Doe qualifies for retroactive application.
By holding that sex offender registration is "punishment" for ex post facto purposes, our supreme court held that the legislature had no authority to order sex offenders to comply with the registration requirements if their underlying offenses predated the enactment of the law. This constitutional ruling falls within Teague's first prong: a new constitutional rule that puts certain kinds of conduct beyond the power of the legislature to regulate through criminal legislation.
The United States Supreme Court addressed this issue in Schriro v. Summerlin, 542 U.S. 348, 124 S.Ct. 2519, 159 L.Ed.2d 442 (2004):
New substantive rules generally apply retroactively [under the Teague test]. This includes decisions that narrow the seope of a criminal statute by interpreting its terms, as well as constitutional determinations that place particular conduct or persons covered by the statute beyond the State's power to punish.... Such rules apply retroactively because they necessarily carry a significant risk that a defendant stands convicted of an act that the law does not make eriminal or faces a punishment that the law cannot impose upon him.
Summerlin, 542 U.S. at 351-52, 124 S.Ct. at 2522-23 (citations, internal quotations, and a footnote omitted).
See United States v. Dashney, 52 F.3d 298, 299 (10th Cir.1995) (holding that, under Teague, a defendant could seek retroactive application of a new judicial interpretation of a criminal statute that altered the generally accepted elements of the crime); United States v. Shelton, 848 F.2d 1485, 1488-1490 (10th Cir.1988) (en bane ) (holding that, under Teague, a defendant was entitled to retroactive application of a substantive (but non-constitutional) decision concerning the reach of a federal statute); Ingber v. Enzor, 841 F.2d 450, 453-54 (2nd Cir.1988) (same); Magnuson v. United States, 861 F.2d 166, 167 (7th Cir.1988) (same); Webster v. Woodford, 369 F.3d 1062, 1068-69 (9th Cir.2004) (holding that, under Teague, a defendant could pursue a due process/ ex post fucto challenge to a new and expanded judicial interpretation of a sentencing statute).
The fact that our supreme court's decision in Doe qualifies for retroactive application under the Teague test means that Doe qualifies for retroactive application under the Judd test-because, in Smart II, 202 P.38d at 1138-39, our supreme court declared that the Judd rule of retroactivity was to be construed as being at least as protective of defendants' rights as the Teague test.
Thus, even under Alaska's current rule of retroactivity (the Judd test), Charles is entitled to retroactive application of Doe. Accordingly, even though the Griffith rule of retroactivity has much to recommend it, we conclude that the resolution of Charles's case does not require the supreme court to decide whether to adopt the Griffith rule.
If the ex post facto clause bors retrospective application of a criminal statute to a defendant, does convicting a defendant for violating that statute constitute "plain error"?
Even though we conclude that Charles is entitled to retroactive application of the Doe decision, the fact remains that Charles did not raise an ex post facto argument during his trial court proceedings, so he is not entitled to relief unless he demonstrates plain error. That is, Charles must show that the ex post facto error in his case was "so prejudicial to the fairness of the *790failure to correct it » 27 proceedings that ... would perpetuate manifest injustice.27
It is undisputed that Charles's underlying sex offense was committed before the sex offender registration act took effect. Thus, Charles stands convicted of violating a criminal statute which, under our state constitution, can not apply to him. This is a manifest injustice.
The record does not suggest that Charles had any tactical reason for deliberately withholding an ex post facto challenge to his prosecution and conviction. Accordingly, we conclude that Charles has shown plain error.
See Mayers v. State, 42 So.3d 33, 44-45 (Miss.App.2010) (holding that a conviction in violation of the ex post facto clause is plain error); State v. Clemons, unpublished, 2011 WL 861847, *4 (Ohio App.2011) (same); People v. Barnes, unpublished, 2010 WL 4970759, *4 (Mich.App.2010) (holding that imposition of increased punishment in violation of the ex post facto clause is plain error); State v. Simmick, 279 Neb. 499, 779 N.W.2d 335, 342 (2010) (same); State v. Houston, unpublished, 2009 WL 130189, *8 (Tenn.Crim.App.2009) (same).
Conclusion
For the reasons explained here, we recommend that the supreme court grant relief to Charles on the basis of the court's decision in Doe. The Clerk of the Appellate Courts is directed to transmit this response to the supreme court.
COATS, Chief Judge, concurring.
. See Charles v. State, Alaska App. Memorandum Opinion No. 1600 (April 27, 1988), 1988 WL 1511427, where this Court affirmed (on direct appeal) Charles's conviction for second-degree sexual abuse of a minor.
. For purposes of this discussion, a criminal conviction is "final" if there is no further possibility of direct appellate review (including discretionary review) of the conviction. See Beard v. Banks, 542 U.S. 406, 411, 124 S.Ct. 2504, 2510, 159 L.Ed.2d 494 (2004); Caspari v. Bohlen, 510 *781U.S. 383, 390, 114 S.Ct. 948, 953, 127 L.Ed.2d 236 (1994); Smart v. State, 146 P.3d 15, 17 (Alaska App.2006), reversed on other grounds in State v. Smart, 202 P.3d 1130 (Alaska 2009).
. AS 11.56.200(c).
. See Dutton v. State, 970 P.2d 925, 931-32 (Alaska App.1999).
. See United States v. Diaz-Diaz, 327 F.3d 410, 412 (5th Cir.2003); State v. Simmick, 279 Neb. 499, 779 N.W.2d 335, 339 (2010); State v. LaFreniere, 342 Mont. 309, 180 P.3d 1161, 1163-64 (2008); Mayers v. State, 42 So.3d 33, 44 (Miss.App.2010); Williams v. State, 507 So.2d 1171, 1171 (Fla.App.1987). But see United States v. Groves, 369 F.3d 1178, 1182 (10th Cir.2004) (holding that the defendant was entitled to challenge his sentence on ex post facto grounds, even though he pleaded guilty, because he did not agree to a specific sentence, and because he reserved the right to appeal any "illegal sentence" that might be imposed).
. Ieppert v. State, 908 SW.2d 217, 220 (Tex.Crim.App.1995).
. Ibid.
. See United States v. Diaz-Diaz, 327 F.3d at 412; State v. Simnick, 779 N.W.2d at 339-340; Mayers v. State, 42 So.3d at 44-45; United States v. Jones, 899 F.2d 1097, 1103 (11th Cir.1990) (holding that failure to raise an ex post facto objection will preclude consideration of the point on appeal absent manifest injustice).
. See Twogood v. State, 223 P.3d 641, 651 (Alaska App.2010); Malutin v. State, 198 P.3d 1177, 1184 (Alaska App.2009); Lockuk v. State, 153 P.3d 1012, 1017-18 (Alaska App.2007) McDole v. State, 121 P.3d 166, 170 (Alaska App.2005); Haag v. State, 117 P.3d 775, 783 (Alaska App.2005).
. See, e.g., Adams v. State, 261 P.3d 758, 773 (Alaska 2011).
. See, e.g., Active v. State, 153 P.3d 355, 367 (Alaska App.2007); Milligrock v. State, 118 P.3d 11, 17 (Alaska App.2005).
. Adams v. State, 261 P.3d 758, 764 (Alaska 2011).
. See Supreme Court Order No. 1759 (both dated and effective July 21, 2011).
. Judd v. State, 482 P.2d 273, 278 (Alaska 1971) (adopting the retroactivity standard formerly employed by the federal courts under Linkletter v. Walker, 381 U.S. 618, 85 S.Ct. 1731, 14 LEd.2d 601 (1965)).
. Griffith, 479 U.S. at 328, 107 S.Ct. at 716.
. See Johnson v. New Jersey, 384 U.S. 719, 732, 86 S.Ct. 1772, 1780, 16 L.Ed.2d 882 (1966); Stovall v. Denno, 388 U.S. at 300, 87 S.Ct. at 1971.
. - Quoting Justice Harlan's dissenting opinion in Desist v. United States, 394 U.S. 244, 258, 89 S.Ct. 1030, 1038, 22 L.Ed.2d 248 (1969).
. See Danforth v. Minnesota, 552 U.S. 264, 128 S.Ct. 1029, 169 L.Ed.2d 859 (2008).
. See Smart v. State (Smart I), 146 P.3d at 29; State v. Smart (Smart II), 202 P.3d at 1148.
. Smart II, 202 P.3d at 1136; "[The United States Supreme Court's decision in] Danforth [v. Minnesota] allows us to apply either the Teague test ... or a state constitutional test[,] so long as the state test is at least as comprehensive as the [Teague] test."
. See Alaska Appellate Rules 507(b) and 512(a)(2). And see footnote 6 of Griffith v. Kentucky, 479 U.S. at 321, 107 S.Ct. at 712.
. Judd v. State, 482 P.2d 273, 278 (Alaska 1971) (adopting the retroactivity standard formerly employed by the federal courts under Linkletter v. Walker, 381 U.S. 618, 85 S.Ct. 1731, 14 L.Ed.2d 601 (1965)).
. See Malloy v. State, 1 P.3d 1266, 1282 (Alaska App.2000).
. 489 U.S. 288, 109 S.Ct. 1060, 103 L.Ed.2d 334 (1989).
. Smart v. State, 146 P.3d at 20-21.
. Teague, 489 U.S. at 307, 109 S.Ct. at 1073 (quoting Justice Harlan's concurring and dissenting opinion in Mackey v. United States, 401 U.S. 667, 692-93, 91 S.Ct. 1171, 1180, 28 L.Ed.2d 404 (1971)).
. Adams v. State, 261 P.3d 758, 764 (Alaska 2011).