April 25, 2014
BYRON E. CHARLES, Petitioner,
STATE OF ALASKA, Respondent.
Petition for Hearing from the Court of Appeals of the State of Alaska No. A-09623, on appeal from the Superior Court of the State of Alaska No. 1KE-05-00765 CR, First Judicial District, Ketchikan, Kevin Miller, Judge.
Tracey Wollenberg, Assistant Public Defender, and Quinlan Steiner, Public Defender, Anchorage, for Petitioner.
Timothy W. Terrell, Assistant Attorney General, Office of Special Prosecutions & Appeals, Anchorage, and Michael C. Geraghty, Attorney General, Juneau, for Respondent.
Before: Fabe, Chief Justice, Winfree, Stowers, and Maassen, Justices, and Eastaugh, Senior Justice. [Bolger, Justice, not participating.]
EASTAUGH, Senior Justice.
Does the 2008 holding in Doe v. State (Doe I) apply to Byron E. Charles, whose 2006 conviction for violating the Alaska Sex Offender Registration Act (ASORA) was still on direct review when he argued for the first time, in reliance on our then-recent Doe I opinion, that applying ASORA to him violated Alaska's ex post facto clause? We conclude that it does. In doing so, we adopt for cases on direct review the federal retroactivity standard announced in Griffith v. Kentucky. We also conclude that Charles's prior failure to raise the ex post facto issue does not bar him from doing so now: Manifest injustice would result if he could not challenge on direct review his conviction for violating a criminal statute that, under our constitution, may not be applied to him.
We therefore reverse the court of appeals's 2007 decision that affirmed Charles's judgment and reverse his 2006 judgment of conviction.
II. FACTS AND PROCEEDINGS
Byron E. Charles was convicted of a sex offense in the 1980s. In 1994 the Alaska Legislature enacted ASORA. The statute was expressly retroactive: It stated that persons convicted of sex offenses were required to register as sex offenders "regardless of whether the conviction occurred before, after, or on the effective date of [ASORA]."  ASORA accordingly required Charles to maintain sex offender registration.
In 2006 Charles was charged with misdemeanor failure to register as a sex offender. At Charles's failure-to-register trial, a Ketchikan police officer testified that he checked Charles's 2005 registration and found that the listed address did not exist. Charles testified that he had written the nonexistent address, but asserted that he likely confused a past address with a family member's address. The trial judge, sitting as the trier of fact, found that Charles was aware of a substantial probability that the address was inaccurate and was thus guilty of failure to register. A judgment of conviction for misdemeanor failure to register as a sex offender was entered against Charles in 2006.
Charles filed a timely appeal in which he argued only that the trial court's finding of guilt was inconsistent with its finding Charles was credible, that the trial court erred in ruling on an evidentiary issue, and that the evidence was not sufficient to sustain the conviction. He did not argue that applying ASORA to him would be an ex post facto violation. The court of appeals affirmed Charles's conviction in 2007.
In 2008 we issued our opinion in Doe I, holding that applying ASORA to Doe violated the Alaska Constitution's ex post facto clause. Charles then filed a timely pro se petition for hearing to this court; relying on Doe I, he argued that his failure-to-register conviction violated the ex post facto clause. Charles had never raised an ex post facto argument in the superior court or in his direct appeal to the court of appeals. The State opposed Charles's petition, arguing that Doe I did not retroactively excuse Charles's failure to comply with ASORA. In light of Doe I's possible impact on Charles's conviction, we remanded to the court of appeals to consider these questions:
1. Assuming Charles is now raising an ex post facto challenge to application of ASORA to him, did Charles waive that argument, and if so, is an ex post facto challenge waivable?
2. Should the principle of direct review retroactivity as described in the federal courts, see Griffith v. Kentucky, 479 U.S. 314 (1987), be adopted as a principle of Alaska law, and if so, under the principle of direct review retroactivity, should Charles's 2006 conviction for failure to register be set aside? 
On remand, Charles (by then represented by counsel) and the State filed supplemental briefs and presented oral arguments. Noting that we had not entered a final order with respect to Charles's petition, and had thus not explicitly either granted or denied the petition, the court of appeals treated the remand as a request for its "input and recommendations on these issues, " not as a request to decide these issues. Its per curiam opinion suggested the following: (1) Ex post facto rights should be intentionally waivable and unintentionally forfeitable; (2) Charles unintentionally forfeited his ex post facto claim, limiting review to plain error; (3) Doe I should retroactively apply to Charles's case under either Alaska's existing retroactivity standard (the Judd standard ) or the federal standard for direct review retroactivity (the Griffith standard ), making it unnecessary to decide whether to adopt the federal standard; and (4) Charles's conviction violated the ex post facto clause and was plain error. Chief Judge Coats concurred but wrote separately to emphasize the benefits of retaining Alaska's existing retroactivity standard.
After the court of appeals issued its per curiam opinion, we asked the parties to file supplemental briefs.
III. STANDARD OF REVIEW
This petition presents only questions of law, to which we apply our independent judgment.
We now resume our consideration of the petition for hearing Charles filed in 2008. Charles was convicted of failing to register as a sex offender under ASORA. In Doe I we ruled that ASORA cannot constitutionally be applied to persons who committed their underlying sex offenses prior to the statute's enactment. As we recognized by entering our order that remanded Charles's case to the court of appeals, there are two potential barriers to applying Doe I to Charles's conviction. First, because Doe I was issued after Charles's 2006 conviction, there is an issue whether Doe I applies to him retroactively. Second, because Charles raised no ex post facto challenge in the trial court or the court of appeals, there is an issue whether Charles should be barred from raising that challenge for the first time in his petition for hearing.
Charles's supplemental briefing asks us to adopt the recommendations of the court of appeals and vacate his conviction without requiring full briefing on the merits. Charles also argues that we should adopt the federal Griffith standard and accordingly hold that Doe I retroactively bars his conviction. He argues alternatively that even if we do not adopt the Griffith standard, Doe I retroactively applies to Charles under the existing Judd standard. He contends that no plain error analysis is necessary, but that, in any event, his conviction was plain error.
In its supplemental brief the State agrees that we should adopt the Griffith standard. But it argues that Charles forfeited his ex post facto claim and that review should be limited to plain error. As to the merits, it refers us to its arguments in State v. Doe A and State v. Stickman. The State's supplemental brief asks us to wait for those cases to be resolved and to require full briefing before deciding whether Charles's conviction should be reversed. As we discuss below in Part IV.C, Doe A and Stickman have since been resolved.
A. We Adopt The Griffith Standard For Direct Review Retroactivity.
To determine whether Doe I retroactively applies to Charles's case, we begin by analyzing the applicable standard for retroactively applying new constitutional rules. Broadly speaking, the choice is between two standards: either full retroactivity or limited ("direct review") retroactivity. As we said in State v. Smart, "[a]pplying a new constitutional principle to criminal cases on collateral review gives the principle 'full retroactivity, ' " and "[a]pplying a new principle to cases on direct review, to cases at the trial level, to any case not yet charged, and to crimes not yet committed gives the principle 'limited retroactivity.' " A case is "final" when "a judgment of conviction has been rendered, the availability of appeal exhausted, and the time for a petition for certiorari elapsed or a petition for certiorari finally denied." Because we issued Doe I before Charles's time to file a petition for hearing had expired, Charles's case was not yet final, and we are thus presented with an issue of limited, or direct review, retroactivity.
In Judd we identified three criteria to determine whether new constitutional rules must apply retroactively: "(a) the purpose to be served by the new standards; (b) the extent of the reliance by law enforcement authorities on the old standards; and (c) the effect on the administration of justice of a retroactive application of the new standards."We adopted those criteria from the federal standard announced in 1965 by the United States Supreme Court in Linkletter v. Walker. In doing so we noted that the criteria "apparently [had] been agreed upon by all authorities." But in Griffith, issued in 1987, the United States Supreme Court abandoned the standard announced in Linkletter for direct review retroactivity and instead instructed that new constitutional rules will automatically apply to cases that are not yet final or are on direct review when the new rule is announced. In State v. Smart, issued after Griffith, we explained that "it may be difficult to defend the use of the Linkletter/Judd standards in cases involving direct review on the state level."
Both Charles and the State ask us to abandon the Judd standard for direct review retroactivity and adopt Griffith. They reason that adopting Griffith will promote fairness and support Alaska's equal protection and due process guarantees by ensuring that similarly situated defendants — those with cases on direct review — are treated similarly. The court of appeals's per curiam opinion suggests that under either the Judd standard or the Griffith standard our holding in Doe I should retroactively bar Charles's conviction. The per curiam opinion therefore concludes that we do not need to decide whether to adopt the Griffith standard.
Our review of the United States Supreme Court's reasons for abandoning the Linkletter criteria persuades us that we should follow suit for direct review retroactivity. The abandonment of Linkletter began in Williams v. United States.There, the Court followed Linkletter and denied retroactive application of a new rule to a case on direct review. But Justice Harlan dissented. He argued that new constitutional rulings should always be applied to cases on direct review; he explained that it was indefensible to "[s]imply fish one case from the stream of appellate review, us[e] it as a vehicle for pronouncing new constitutional standards, and then permit a stream of similar cases subsequently to flow by unaffected by that new rule." His dissent steadily gained acceptance. In 1982 in United States v. Johnson, the Supreme Court adopted it in place of Linkletter for cases on direct review, although the Court retained an exception if the new constitutional ruling was a "clear break" from the past rule. Five years later, in Griffith, the Supreme Court dispensed with the "clear break" exception and thus fully adopted the standard Justice Harlan had proposed for cases on direct review.
In Griffith the Supreme Court explained that failing to apply new constitutional rules to cases on direct review violated two norms of constitutional adjudication: "the principle that this Court does not disregard current law" and the principle of "treating similarly situated defendants the same."  With regard to the first principle, the Court explained that the "integrity of judicial review" requires the Court to resolve cases before it on direct review "in light of [the Court's] best understanding of governing constitutional principles." With regard to the second principle, the Court explained in subsequent decisions that Linkletter's case-by-case weighing of interests "proved difficult to apply in a consistent, coherent way, " leading to "strikingly divergent results" and "the disparate treatment of similarly situated defendants on direct review ." In G riffith the Co urt therefore "refused to continue to tolerate the inequity that resulted from not applying new rules retroactively to defendants whose cases had not yet become final."
Because Judd adopted Linkletter's case-by-case analysis,  the current Alaska standard is susceptible to the same disparate application of new constitutional rulings: The defendant in the case in which the ruling is announced will receive the benefit of the ruling, but other defendants with cases still pending before us may be denied application of our best understanding of governing constitutional principles. We now have the opportunity to decide whether to follow the same path taken by the United States Supreme Court and abandon the Linkletter/Judd criteria for direct review retroactivity.
We overrule a prior decision only when we are "clearly convinced that (1) a decision was originally erroneous or is no longer sound because of changed conditions; and (2) more good than harm would result from overruling it."
In Judd we explained that retroactivity standards involve "a value judgment" and that we gave substantial weight to public confidence in the judiciary.W e reasoned that because police, prosecuting authorities, and the public rely on previous statements of the law, "the application of a new interpretation to past conduct which was accepted by previous judicial decisions leads us to confusion and a hesitancy to accept any theory except one of gamesmanship with corresponding disrespect for our whole system of laws." We relied on what we there termed "the practical problems" of public and law enforcement reliance on previous statements of law as justification for adopting two of the Linkletter criteria — the extent of reliance by law enforcement on past rules and retroactivity's effect on the administration of justice. But we also recognized a countervailing interest: Denying retroactive application of a new ruling to cases pending on direct review leads to disparate application of legal rules and "in a particular case the court may appear to countenance illegal acts on the behalf of the police." Indeed, this interest in preventing unfair or unconstitutional prosecutions is enshrined in the remaining Linkletter criterion — the purpose to be served by the new rule. By weighing the purpose of the new rule against law enforcement reliance and administrative burdens, the Judd balancing test enables, but does not guarantee, the application of new constitutional rules to cases pending on direct review. In this way, the Judd balancing test can ameliorate some concern with courts having a hand in the "dirty business" of facilitating prosecutions stemming from unconstitutional laws or procedures. And we have held that if the purpose of a new rule is to ensure the fairness of a trial, that purpose always outweighs the countervailing interests and allows for retroactive application to cases on direct review. Nonetheless, if a rule is not clearly related to the fairness of a trial — i.e., "where the purpose . . . is not to minimize arbitrary or unreliable fact findings" — Judd still potentially allows the inequity of selectively applying a new rule to the party before us while denying its application to other parties with cases currently pending on direct review.
For the reasons the United States Supreme Court discussed in Griffith, we now conclude that disregarding our best understanding of constitutional principles and disparately applying legal rules to cases on direct review undermines the integrity of judicial review and violates the principle of treating similarly situated defendants the same. We therefore hold that Judd was erroneous to the extent it potentially limited the application of new constitutional rules to criminal cases pending on direct review.
We also conclude that more good than harm will come from adopting Griffith's bright-line standard for direct review retroactivity. Applying our best understanding of constitutional law to all defendants on direct review does more good than allowing disparate treatment merely because of law enforcement reliance or administrative burdens. By requiring the application of our best understanding of the constitution to all cases pending before us, Griffith leads to more rational results than Judd. And in guaranteeing similar treatment of similarly situated defendants, Griffith is more fair than Judd. Furthermore, adopting Griffith will not result in overwhelming administrative burdens. Indeed, abandoning Judd and adopting Griffith for cases on direct review does not shift the law far from how we have applied Judd in practice: to require direct review retroactivity "where a new rule serves to ensure defendants a fair trial." Although Griffith potentially may result in more reversals, it removes the uncertainty of litigating whether a new rule relates to the fairness of a trial and guarantees uniform application of new constitutional rules to similarly situated defendants. Furthermore, in adopting Griffith we will be able to draw from an available body of federal law . A nd our state-law retroactivity analysis will align more closely with the federal-law analysis that guarantees that we satisfy minimum federal standards.
Because we conclude that in retrospect it was error to adopt the Linkletter criteria for cases on direct review and that adopting Griffith for A laska will do more good than harm, we overrule Judd and adopt Griffith as the standard for direct review retroactivity in Alaska.
We decline to follow the suggestion of the court of appeals that we defer this determination. The per curiam opinion reasoned that Doe I would retroactively apply to Charles's case under the Judd standard and that "even though the Griffith rule of retroactivity has much to recommend it, " this court does not need to rule on whether to adopt it. It is correct that Judd would appear to allow retroactive application of Doe I: Holding the application of ASORA unconstitutional to persons in Charles's circumstances relates to the fairness of the prosecution. Retroactively applying that factor to cases on direct review would therefore fairly extend our current notion that rules that ensure a fair trial or relate to the integrity of the verdict should be given retroactive effect.  Nonetheless, it is our task to make the necessary policy decisions and value judgments in discerning retroactivity standards. We perceive no plausible reason for delaying to another day the question of whether to adopt Griffith. Both Charles and the State have persuasively recited the reasons for adopting Griffith. And the Supreme Court's reasons for limiting and ultimately abandoning Linkletter are convincing. We therefore conclude that litigants on direct review should receive the benefit of Griffith's clear and fair retroactivity standard.
We now turn to applying to Charles's case the retroactivity standard we adopt today — that all new constitutional rules apply to similar cases that are pending on direct review or not yet final. We announced our holding in Doe I while Charles could file a timely petition for hearing, and he ultimately filed his petition raising the ex post facto issue within the time required for filing his petition. Because his case was therefore on direct review, he retroactively receives the benefit of Doe I. The issue remains, however, whether Charles waived or forfeited his Doe I argument by failing to raise the issue at trial or on direct appeal to the court of appeals.
B. Charles Did Not Waive His Ex Post Facto Challenge.
On remand, we asked the court of appeals to consider whether an ex post facto challenge is waivable and whether Charles had waived his challenge. The court of appeals noted that nothing in the record suggests Charles knowingly decided not to pursue an ex post facto challenge at trial. The court of appeals therefore explained that the specific issue is whether ex post facto challenges may be unintentionally forfeited.It noted that "[s]everal federal and state courts have ruled that [ex post facto] protections can be forfeited by failing to assert them, " meaning the party must then demonstrate plain error. In comparison, the Texas Court of Criminal Appeals has ruled that because ex post facto challenges attack the legislature's authority to criminalize conduct, "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.' "
Charles argues that we should follow Texas and hold that his ex post facto challenge — an attack on the constitutionality of ASORA — is jurisdictional in character, and can thus be raised at any time.
Because there is no indication that Charles intentionally waived an ex post facto challenge, we do not need to consider whether the court of appeals was correct in stating that a criminal defendant may intentionally waive an ex post facto challenge.
And because Charles did not raise his ex post facto challenge before the district court or the court of appeals, we would ordinarily review the challenge only for plain error.  As the court of appeals noted, "[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."  But that explanation of plain error does not adequately explain how to deal with a party's post-trial attempt to invoke a new constitutional rule that had not yet been announced when the case was tried and that was therefore not available to the trial court. Any error in failing to conclude that ASORA violated the ex post facto clause could not fairly be called obvious to any judge of a court in which the issue might have been raised: We had not yet issued our ruling in Doe I when Charles's failure-to-register case was tried and appealed to the court of appeals.
The court of appeals proposed this special, "retrospective" plain error rule when a new constitutional rule is announced: "The question is whether, in light of the new constitutional rule, we can now see that there was an obvious flaw in the lower court proceedings, and that failure to correct this flaw would perpetuate manifest injustice."But we think this formulation, although helpful, adds avoidable complexity to the ultimate inquiry.
We therefore turn to the core of our plain error analysis. In Adams v. State, we stated: "We have described plain error as involv[ing] such egregious conduct as to undermine the fundamental fairness of the trial and contribute to a miscarriage of justice." We agree with that description. Therefore, in deciding whether it was plain error not to apply a not-yet-announced constitutional rule, instead of focusing on whether the error was hypothetically obvious, we ask whether the error was "so prejudicial to the fairness of the proceedings that . . . failure to correct it would perpetuate manifest injustice."
The court of appeals concluded that because Doe I ruled that ASORA "violates Alaska's ex post facto clause, " and because it is undisputed that Charles committed his sex offense prior to ASORA's enactment, "Charles stands convicted of violating a criminal statute which, under our state constitution, can not apply to him."W e agree with that conclusion. W e also conclude that permitting Charles to be convicted of violating a criminal statute that cannot constitutionally be applied to him would result in manifest injustice. Charles may therefore rely on Doe I even though he failed to raise an ex post facto challenge at trial. His conviction must be set aside.
C. We Decline To Allow Additional Briefing Before Granting Relief.
The State's initial opposition to Charles's petition for hearing argued that we should deny the petition, because, even though Doe I applied to the class of sex offenders that included Charles, it was not unconstitutional to charge him with failing to comply with ASORA before we issued our Doe I decision in 2008.
When we ordered supplemental briefing after the court of appeals issued its per curiam opinion on remand, Charles's supplemental brief argued that any conviction based on a registration duty that could not be constitutionally applied to Charles was invalid. It asked us to vacate Charles's conviction of failing to register.
In response, the State's supplemental brief argued that we should not reverse Charles's conviction without full briefing on the merits. It also argued — summarizing two arguments it had made in its then-pending appeal in State v. Doe — that Doe I had been wrongly decided and was not binding precedent. The State's supplemental brief acknowledged that under Doe I, "prospective enforcement of a sentence stemming from such a failure-to-register conviction must cease." But it argued — summarizing arguments it had made in State v. Stickman,  an appeal then pending in the court of appeals — that offenders should not have their convictions declared void ab initio or expunged. The State's supplemental brief concluded that we "should decline to summarily reverse and vacate Charles'[s] conviction until this court decides [Doe A], . . . and, if necessary, Alaska's appellate courts resolve the state's additional arguments in State v. Stickman."
Thus, the State's supplemental brief in this case gave two reasons for asking us to delay ruling: First, that Doe I was wrongly decided and was not binding precedent. Second, that the State's arguments in State v. Stickman should be resolved. Neither reason persuades us that we should call for additional briefing before ruling on Charles's request that we order that his conviction be vacated.
We issued our decision in Doe A on March 13, 2013, and modified it when we denied rehearing on April 10, 2013. Our Doe A decision expressly held that Doe I was correctly decided and is binding precedent. We have thus considered, and rejected, the two Doe A-related arguments the State's supplemental brief asked us to consider before granting relief to Charles.
The State's appeal in Stickman elaborated on the same arguments the State raised when it first opposed Charles's petition for hearing. In Stickman, the State appealed an order granting post-conviction relief to Donald Stickman after he was convicted of failure to register for a sex offense that pre-dated ASORA. The State's briefs contended there that Stickman had an enforceable duty to comply with ASORA until that statute was declared ex post facto. In support of that contention, the State cited Jacko v. State and Clark v. State. In Jacko the court of appeals rejected George Jacko's argument that a charge for violating a restraining order should be dismissed because the order was factually unjustified. The court of appeals reasoned in Jacko that a person must obey a restraining order until it is vacated through proper procedures.It explained that it would encourage disobedience and violence to rule that "a person may flout a court order with impunity if it later turns out that the order was illegal." In Clark, the court of appeals affirmed a felon-in-possession-of-a-firearm conviction even though the underlying felony conviction was later reversed.
Responding to the State's appeal, Stickman asserted that Doe I abrogated the legislature's authority to require Stickman to register, making any conviction based on failure to register void. Stickman distinguished Jacko and Clark, arguing that court rulings had invalidated the predicate factual circumstances for the restraining order in Jacko and the prohibition on firearm possession in Clark. In comparison, Stickman argued that Doe I's ex post facto holding instead pertained to the constitutionality of the very duty to register. Stickman argued that applying Jacko and Clark to an ex post facto challenge and ruling that a person could be convicted of failing to comply with an unconstitutional statute would undermine a defendant's ability to raise constitutional challenges to the statute under which he or she was convicted.
The State did not file a reply to Stickman's appellee's brief; instead, after obtaining two extensions in which to file a reply, it moved to dismiss its appeal. The court of appeals granted the State's motion to dismiss on May 23, 2013.
Having reviewed the briefs filed by the State and Stickman in File No. A-10441, we are unpersuaded that it would be beneficial to give the State any additional opportunity to argue now that Charles could be charged with violating a statute which we later held could not be constitutionally applied to persons like Charles. First, the State already had an opportunity, which it exercised, to argue that issue when it first opposed Charles's petition for hearing. Second, the State's briefing on that issue in Stickman is not so compelling that we think further briefing might change the outcome of Charles's petition. Third, the State voluntarily dismissed its appeal in Stickman, waiving its opportunity to present the merits of that argument to the court of appeals. It thus waived an opportunity to have, in the words of the State's supplemental brief, an Alaska appellate court "resolve the state's additional arguments in [Stickman]." We do no t decide here whether Jacko and Clark or the analysis underlying those decisions could be the basis for upholding a person's conviction for failing to comply with a criminal statute later declared to be unenforceable against that person for ex post facto reasons.
We merely decide that, given the procedural circumstances in this case and in Stickman, there is no reason to allow additional briefing before we decide Charles's petition.
For these reasons we GRANT Charles's petition for hearing, REVERSE the 2007 decision of the court of appeals that affirmed Charles's judgment, and REVERSE Charles's 2006 judgment of conviction.