Opinion
22SC589
06-10-2024
The People of the State of Colorado, Petitioner v. Charles James Crabtree, Respondent
Attorneys for Petitioner: Philip J. Weiser, Attorney General Carmen Moraleda, Senior Assistant Attorney General Denver, Colorado Attorney for Respondent: Mallika L. Magner Crested Butte, Colorado
AS MODIFIED JULY 1, 2024
Certiorari to the Colorado Court of Appeals Court of Appeals Case No. 20CA629
Attorneys for Petitioner: Philip J. Weiser, Attorney General Carmen Moraleda, Senior Assistant Attorney General Denver, Colorado
Attorney for Respondent: Mallika L. Magner Crested Butte, Colorado
JUSTICE SAMOUR delivered the Opinion of the Court, in which CHIEF JUSTICE BOATRIGHT, JUSTICE MARQUEZ, JUSTICE HART, and JUSTICE BERKENKOTTER joined. JUSTICE HOOD concurred in part and dissented in part. JUSTICE GABRIEL dissented.
SAMOUR, JUSTICE
¶1 Stare decisis may be likened to a traffic sign prohibiting U-turns except in very limited and well-defined circumstances. This judge-made doctrine prevents courts from doing a volte-face on preexisting rules of law unless there are sound reasons to do so-i.e., we must be clearly convinced that (1) the preexisting rule was originally erroneous or is no longer sound because of changing conditions and (2) more good than harm will come from departing from that rule. Not surprisingly, adhering to stare decisis principles, courts are quite reluctant to depart from preexisting rules. Today, however, we make a rare jurisprudential U-turn because we conclude, on the first issue we confront, that there are sound reasons to change course on an aspect of our recent holding in Linnebur v. People, 2020 CO 79M, ¶ 2, 476 P.3d 734, 735. But having made that U-turn, we decline to make a second. Instead, as it relates to the second issue before us, we stay the course set by our case law.
Here are the two issues we agreed to review:
1. Whether the court of appeals erred in using a plain-error standard in light of Linnebur v. People, 2020 CO 79M, 476 P.3d 734.
2. Whether the court of appeals erred in holding that under plain error, it does not matter whether the law was settled at the time the error occurred, so long as the error is "obvious" at the time of appeal, in contravention of this court's holding in Scott v. People, 2017 CO 16[, 390 P.3d 832].
The People raised only the second issue in their petition. We added the first issue nostra sponte.
¶2 In Linnebur, we held that the fact of a defendant's relevant convictions in a felony DUI trial is an element of the offense that "must be proved to the jury beyond a reasonable doubt," not a sentence enhancer that "a judge may find by a preponderance of the evidence." ¶ 2, 476 P.3d at 735. Because the trial court there treated the fact of Linnebur's relevant convictions as a sentence enhancer, not as an element of felony DUI, the prosecution didn't have to prove that fact to the jury beyond a reasonable doubt. Id. at ¶¶ 4-5, 476 P.3d at 735-36. And because Linnebur was subsequently sentenced for felony DUI, a crime that differed from the one on which the jury's guilty verdict was based (misdemeanor DUI), we concluded that his felony DUI conviction and corresponding sentence had to be reversed. Id. at ¶ 32, 476 P.3d at 741. In so doing, we impliedly, but necessarily, determined that the error that occurred was a structural one requiring automatic reversal. Id. at ¶¶ 32-33, 476 P.3d at 741-42.
DUI and DWAI are ordinarily misdemeanors, but they are elevated to class four felonies if the violation occurs after three prior convictions for certain alcohol-related driving offenses. § 42-4-1301(1)(a), (b), C.R.S. (2023).
Although Linnebur was actually convicted of and sentenced for felony DWAI, we sometimes referred to his conviction and sentence as being for felony DUI, and our holding was phrased accordingly. See Linnebur, ¶ 2, 476 P.3d at 735. For the sake of consistency and to avoid confusion, here, we refer to Linnebur's conviction and sentence as being for felony DUI.
¶3 The error that occurred in Linnebur's trial has reared its ugly head again. But we now make a U-turn and hold that a Linnebur error is not structural and is, instead, a trial error. And because this trial error was not preserved here, we review for plain error. Therefore, the division of the court of appeals below correctly applied the plain error standard of reversal, even if it did so without explanation or justification for deviating from the remedy meted out in Linnebur. See People v. Crabtree, 2022 COA 73, ¶¶ 41-52, 519 P.3d 415, 423-25.
¶4 But the division went on to hold that where, as here, a trial court's decision was clearly correct at the time it was made but becomes plainly erroneous due to an intervening authoritative legal decision by the time of appeal, it suffices that the error is obvious at the time of appellate consideration. Id. at ¶¶ 47-51, 519 P.3d at 424-25. The division anchored its holding to jurisprudence from the United States Supreme Court. Id. We, however, have employed a different analytical framework in applying our plain error rule, Crim. P. 52(b), than the one the Supreme Court uses in applying the federal plain error rule, Fed. R. Crim. P. 52(b). As relevant here, we have repeatedly said that the plainness of an error for purposes of plain error review must be judged at the time the error is made.
Crim. P. 52(b) permits Colorado's appellate courts to review for plain error if (1) there is an error, (2) that is plain, and (3) that affects the defendant's substantial rights. When an unpreserved error meets all three criteria, relief is mandatory. See Crim. P. 52(b).
¶5 The defendant, Charles James Crabtree, nevertheless nudges us to follow in the division's footsteps and cherry-pick one part of the Supreme Court's plain error methodology-the time-of-appeal rule adopted in Henderson v. United States, 568 U.S. 266, 273 (2013), for determining the plainness of an unpreserved error.In Henderson, the Court held that, so long as an error is obvious at the time of appeal, it doesn't matter whether the law was settled or unsettled at the time of trial. Id. at 273-75.
Under Supreme Court jurisprudence, an appellate court may review for plain error pursuant to Fed. R. Crim. P. 52(b) if (1) there is an error; (2) that is plain; (3) that affects the defendant's substantial rights; and (4) that seriously affects the fairness, integrity, or public reputation of judicial proceedings. See United States v. Olano, 507 U.S. 725, 732-36 (1993). Relief under Fed. R. Crim. P. 52(b) is discretionary-in deciding whether to grant relief, federal courts exercise their discretion by applying what's commonly known as Olano's fourth prong (whether the unpreserved error seriously affects the fairness, integrity, or public reputation of judicial proceedings). See id. at 736.
¶6 But on this point, we stand steadfastly by our case law and decline to follow Henderson for three reasons: (1) the rationale for embracing the time-of-appeal rule was partially based on a concern that is irrelevant in Colorado state courts (the unfairness of treating defendants in different federal circuits unevenly based on the state of the law in each circuit at the time an error occurs), see id.; (2) unlike the federal system, Colorado has a postconviction vehicle to address a "significant change in the law" during the pendency of a direct appeal, see Crim. P. 35(c)(1); and (3) the Supreme Court allayed concerns about the prospect of opening the "'plain error' floodgates" by relying on, among other things, the screening criterion that's part of the federal plain error standard but absent from Colorado's plain error standard (Olano's fourth prong), Henderson, 568 U.S. at 276. As these differences underscore, plucking a single piece of the federal plain error standard while disregarding the rest of the standard is fraught with peril. Because our time-of-trial rule was not originally erroneous and continues to be sound, there is no basis to stray from principles of stare decisis and switch to a time-of-appeal rule.
¶7 What about replacing our plain error standard under Crim. P. 52(b) in its entirety with the Supreme Court's plain error standard under Fed. R. Crim. P. 52(b)? After all, in the past, we've endeavored to construe our state rules in conformity with the federal courts' interpretations of our rules' federal counterparts-at least where, as here, the rules in question are substantively similar. Indeed, this is avowedly our general preference. We carefully considered that option here. In the end, however, we heed the wisdom of the aphorism "if it ain't broke, don't fix it." Our plain error standard has served us well for many years, and neither the parties in this case nor the wider bar have expressed interest in the federal approach, much less identified deficiencies warranting a complete rebuild of our methodology. Such an overhaul would inevitably be terribly disruptive-unnecessarily so given the postconviction remedy available under Crim. P. 35(c)(1).
¶8 Looking at the unpreserved trial error here, we agree with the parties that it was not obvious at the time it occurred. Therefore, we conclude that it cannot be considered plain and that Crabtree is not entitled to relief under Crim. P. 52(b). Accordingly, we reverse the division's judgment and remand this case with instructions to return it to the trial court for reinstatement of Crabtree's felony DUI conviction and sentence.
I. Facts and Procedural History
¶9 A tow truck driver called the police to report that the driver of a car he was intending to tow seemed to be intoxicated. An officer responded and found Crabtree sitting in the front seat of a running car drinking a beer. Crabtree had bloodshot, watery eyes and struggled to understand what the officer said to him. As Crabtree exited the car at the officer's request, he could not maintain his balance. The officer smelled alcohol on Crabtree's breath, and Crabtree admitted he'd consumed an alcoholic beverage earlier in the day. Two other officers responded later, and they confirmed that Crabtree had bloodshot, watery eyes and that his breath smelled of alcohol. Crabtree refused to perform roadside sobriety tests and declined to take a chemical test to measure the percentage of alcohol in his blood stream.
¶10 The People subsequently charged Crabtree with felony DUI-three or more relevant convictions (i.e., convictions "arising out of separate and distinct criminal episodes" for DUI, DUI per se, DWAI, vehicular homicide, vehicular assault, or "any combination thereof"). See § 42-4-1301(1)(a), C.R.S. (2023). At trial, the jury was asked to decide whether Crabtree was guilty of DUI. But the jury was not told of Crabtree's relevant prior convictions and was not asked to determine whether he had incurred those convictions. Instead, in a separate hearing held after the jury found Crabtree guilty of DUI, the trial judge found that the People had proved by a preponderance of the evidence that Crabtree had been convicted of the four alcohol-related driving offenses listed in the complaint. The judge's finding elevated the jury's guilty verdict from a misdemeanor to a class four felony. See id.
¶11 At the time of Crabtree's trial, there was only one published appellate decision in Colorado addressing whether the fact of a defendant's relevant convictions in a felony DUI case is an element of the offense that must be proved to the jury beyond a reasonable doubt. See People v. Gwinn, 2018 COA 130, ¶¶ 39-56, 428 P.3d 727, 736-39. In Gwinn, a division of the court of appeals answered the question in the negative, concluding that this fact was merely a sentence enhancer that could be proved to a judge by a preponderance of the evidence. Id. at ¶ 39, 428 P.3d at 736. The decision in Gwinn explains why the trial court proceeded as it did here without any objection from Crabtree.
¶12 Then, ten months after the trial in this case, another division of the court of appeals reached the opposite conclusion. See People v. Viburg, 2020 COA 8M, ¶ 28, 477 P.3d 746, 751 ("Viburg I") ("[W]e conclude that the fact of a prior conviction is an essential element of felony DUI that must be proved to a jury beyond a reasonable doubt."). And ten months after that, our court resolved the conflict between Gwinn and Viburg I, landing firmly alongside Viburg I in the element camp. See Linnebur, ¶ 31, 476 P.3d at 741.
¶13 We announced our decision in Linnebur while Crabtree's case was pending in the court of appeals. Availing himself of the new rule established by Linnebur, Crabtree argued to the division below that the trial court had erred in failing to treat the fact of his relevant convictions as an element of felony DUI. Unsurprisingly, the division agreed with him. Crabtree, ¶ 41, 519 P.3d at 423. But, despite our treatment of the same error in Linnebur as structural, the division declined to automatically reverse. Instead, after acknowledging that the issue was not preserved because Crabtree did not object at trial, the division reviewed for plain error to determine whether reversal was warranted. Id. at ¶¶ 41-53, 519 P.3d at 423-25. The division did not explain why it diverged from the remedy meted out in Linnebur.
¶14 In applying plain error review, the division took up plainness first. It noted that at the time of Crabtree's trial, Gwinn reigned supreme because neither the court of appeals' decision in Viburg I nor our decision in Linnebur had yet seen the light of day. Id. at ¶¶ 38-40, 519 P.3d at 423. The division then inquired whether an error may still count as plain if the erroneous nature of the challenged ruling becomes obvious only at the time of appeal. Id. at ¶ 41, 519 P.3d at 423-24. As the division framed the issue, "may an error that is not 'obvious' at the time of trial become so while the case is on appeal, entitling a defendant to the benefit of a change in law?" Id.
¶15 Borrowing heavily from another division's unpublished opinion, id. at ¶ 42 n.4, 519 P.3d at 424 n.4 (citing People v. Houghton, No. 19CA2192 (Jan. 27, 2002)), the division answered "yes," id. at ¶ 42, 519 P.3d at 424. But the division didn't stop at "yes." It added that this was a foregone conclusion because other divisions had previously answered the question the same way based on the Supreme Court's decision in Johnson v. United States, 520 U.S. 461, 467-68 (1997). Crabtree, ¶¶ 44-47, 519 P.3d at 424.
¶16 In Johnson, the Court held that, under Fed. R. Crim. P. 52(b), "where the law at the time of trial was settled and clearly contrary to the law at the time of appeal[,] it is enough that an error be 'plain' at the time of appellate consideration." 520 U.S. at 468. According to the division, at least three other divisions had beaten it to the punch and adopted the ruling in Johnson already. Crabtree, ¶¶ 44-46, 519 P.3d at 424 (first citing People v. O'Connell, 134 P.3d 460, 464-65 (Colo.App. 2005) (noting that Johnson established that plain error occurs where the law at the time of trial was settled and clearly contrary to the law at the time of appeal, but that whether the same rule applies where the law was unsettled at the time of trial remained an open question); then citing People v. Versteeg, 165 P.3d 760, 767 (Colo.App. 2006) (applying Johnson in a postconviction proceeding where the law was "settled" at the time of the error and the error was not plain until the trial court considered the defendant's postconviction claim); and then citing People v. Moore, 321 P.3d 510, 513-14 (Colo.App. 2010) (presupposing that Johnson governs where the law is settled at the time of trial, but reasoning that where the law is unsettled at the time of trial, a plain error analysis should be conducted using the law in effect at the time of trial)). Thus, determined the division, the rule in Colorado is clear: "Johnson governs plain error review . . . when the applicable law was settled at the time of trial but not when the applicable law was not settled at the time of trial." Id. at ¶ 47, 519 P.3d at 424.
¶17 Based on Gwinn, the division viewed the relevant law as settled at the time of Crabtree's trial-to quote the division, "it was settled law that [Crabtree] was not entitled to have his prior convictions for alcohol-related driving offenses tried to a jury beyond a reasonable doubt." Id. at ¶ 50, 519 P.3d at 425. And because Gwinn was clearly contrary to the law at the time of appeal, see Linnebur, ¶ 31, 476 P.3d at 741, the division declared that "this case is a Johnson case." Crabtree, ¶ 50, 519 P.3d at 425. Then, applying Johnson, the division concluded that the error here was obvious and that Crabtree was "able to benefit from the change in law applicable on appeal." Id.
¶18 The division was unmoved by the People's reliance on Scott v. People, 2017 CO 16, ¶ 16, 390 P.3d 832, 835, abrogated on other grounds by Whiteaker v. People, 2024 CO 25, ¶ 25, 547 P.3d 1122, 1128, where we observed that, to qualify as plain error, an error must generally be so obvious at the time of trial that the judge should be able to avoid it without the benefit of an objection. Crabtree, ¶ 51, 519 P.3d at 425. For two reasons, the division didn't feel constrained by Scott: (1) it didn't read Scott as addressing the issue squarely; and (2) on multiple occasions following Scott, our court agreed to review, but ultimately had no occasion to decide, whether to adopt Henderson, which expanded the rule in Johnson by holding that, so long as an error is obvious at the time of appeal, it doesn't matter whether the law was settled or unsettled at the time of trial. Id.
¶19 As it relates to the second reason, the division pointed out that in one of the cases in which we agreed to consider whether to adopt Henderson's time-of-appeal rule, we said that we'd "continue to leave that question open." Id. (quoting Campbell v. People, 2020 CO 49, ¶ 41, 464 P.3d 759, 767 (declining to decide whether to adopt Henderson because the error in question was not obvious even on appeal)). The division apparently understood this statement to refer not to whether we should replace the time-of-trial rule we applied in Scott with the time-of-appeal rule the Supreme Court applied in Henderson, but to whether Colorado should adopt for the first time a time-of-trial rule or a time-of-appeal rule to determine the plainness of an error. Id. And, believing that question to be "open," the division doubled down on its conclusion that Scott was not on point and that it was free to endorse Henderson to fill the purported void in our plain error law. Id. With Henderson as its escort, the division held that, so long as an error is obvious at the time of appeal, it is plain for purposes of Crim. P. 52(b). Id. at ¶¶ 41-42, 51, 519 P.3d at 423-25.
This is a nuanced, but important, distinction. It's one thing to consider whether to change an approach already in use with a different one; it's quite another to consider whether to adopt a new approach where there is a blank slate and no approach currently in use. The division read our statement in Campbell to mean the latter.
¶20 The People filed a petition for certiorari, which we granted. And today, after multiple false starts in cases in which we ended up leaving the question unresolved, we decide once and for all whether to adopt Henderson's time-of-appeal rule for purposes of plain error review.
II. Standard of Review
¶21 Whether the division correctly reviewed for plain error is a question of law. Dep't of Corr., Denver Reception &Diagnostic Ctr. v. Stiles, 2020 CO 90M, ¶ 31, 477 P.3d 709, 716. So is whether the division correctly held that, for purposes of plain error review under Crim. P. 52(b), it doesn't matter whether the law was settled or unsettled at the time the error occurred, so long as the error is obvious at the time of appeal. Id. "We review questions of law de novo." People v. Rigsby, 2020 CO 74, ¶ 11, 471 P.3d 1068, 1072.
III. Analysis A. Did the Division Incorrectly Review for Plain Error?
¶22 Errors. They happen-including in criminal jury trials. After all, "to err is human," as the famous poet, Alexander Pope, said.
¶23 When a trial error does happen, what is the standard of reversal? More to the point, when, as here, a trial court errs by failing to require the People to prove to the jury beyond a reasonable doubt the element of a defendant's relevant convictions in a felony DUI case, do we automatically reverse the judgment and sentence if the jury returns a guilty verdict?
¶24 We have previously discussed the law governing the nature of errors that occur in the trial process and the possible remedies for such errors. See James v. People, 2018 CO 72, ¶¶ 11-15, 426 P.3d 336, 338-39. Colorado has largely come to accept the "structural error/trial error dichotomy" adopted by the Supreme Court. Id. at ¶ 15, 426 P.3d at 339.
¶25 James treaded the trail blazed by People v. Novotny, 2014 CO 18, ¶ 2, 320 P.3d 1194, 1196, where, overruling prior cases, we concluded that "reversal of a criminal conviction for other than structural error, in the absence of an express legislative mandate or an appropriate case specific, outcome-determinative analysis, can no longer be sustained." Under Novotny, with one exception (when expressly mandated by the legislature), structural errors are now the only errors in Colorado that require automatic reversal (i.e., reversal in all circumstances without a casespecific, outcome-determinative analysis). Id.
Relying on United States v. Cotton, 535 U.S. 625, 632 (2002); Johnson, 520 U.S. at 469; and federal circuit court cases, the People contend that unpreserved structural errors do not require automatic reversal and should instead be subject to plain error review. We need not, and thus do not, reach this question.
¶26 The "structural error/trial error dichotomy" to which we "firmly adhere[d]" in Novotny "has greatly narrowed" the class of errors to which automatic reversal applies. Id. at ¶ 21, 320 P.3d at 1201. This limited category of errors has been described by the Supreme Court as encompassing three general categories of errors: (1) errors concerning rights that protect an interest other than the defendant's interest in not being erroneously convicted; (2) errors the effects of which are too difficult to measure because they are necessarily unquantifiable and indeterminate; and (3) errors that can be deemed to always result in fundamental unfairness. James, ¶ 15, 426 P.3d at 339-40 (first citing Weaver v. Massachusetts, 582 U.S. 286, 295-96 (2017); and then citing United States v. Gonzalez-Lopez, 548 U.S. 140, 148 (2006)); see also Novotny, ¶ 21, 320 P.3d at 1201 (describing the limited class of errors to which the "structural error" designation applies as "errors that infect the entire trial process and necessarily render a trial fundamentally unfair-rather than simply errors in the trial process itself"). The short list of errors that have been recognized as structural includes the complete deprivation of counsel, a trial before a biased judge, the unlawful exclusion of people of the defendant's race from a grand jury, the denial of the right to selfrepresentation, and the denial of the right to a public trial. Hagos v. People, 2012 CO 63, ¶ 10, 288 P.3d 116, 119.
¶27 As for nonstructural errors (i.e., trial errors), several standards of reversal exist because those errors are not subject to automatic reversal. For preserved trial errors, whether constitutional or unconstitutional, reversal hinges on an appropriate case-specific, outcome-determinative analysis (i.e., constitutional or nonconstitutional harmless error review). Novotny, ¶ 27, 320 P.3d at 1203; see also Hagos, ¶¶ 11-12, 288 P.3d at 119 (same). And for errors that have not been preserved by objection (whether constitutional or unconstitutional), the case- specific, outcome-determinative standard of reversal we apply is plain error review. Hagos, ¶ 14, 288 P.3d at 120.
We identified an additional category of errors in Hagos. For certain types of claims, such as claims for ineffective assistance of counsel, "the error [must] impair the reliability of the judgment of conviction to a greater degree than the harmless error standard requires." Hagos, ¶ 13, 288 P.3d at 119-20. These claims, by their nature, must show prejudice to the trial as a whole. Id., 288 P.3d at 119.
¶28 The remedy we provided Linnebur for the error that occurred in his trial (the same error that occurred in Crabtree's trial) was automatic reversal of the judgment of conviction and sentence (without undertaking an individualized outcome-determinative analysis). Linnebur, ¶ 32, 476 P.3d at 741. Hence, although we didn't label the error as structural, we treated it as such. Id. Citing to Medina v. People, 163 P.3d 1136, 1141-42 (Colo. 2007), we concluded that, "[b]ecause Linnebur was sentenced for a crime different from the one on which the jury's verdict was based, his conviction of felony DUI and sentence must be reversed." Linnebur, ¶ 32, 476 P.3d at 741. But as we recently recognized in another case involving a Linnebur error, the error we addressed in Medina is distinguishable from a Linnebur error. See People v. Viburg, 2021 CO 81M, ¶¶ 26-27, 500 P.3d 1123, 1129-30 ("Viburg II").
¶29 In Medina, the prosecution inadvertently failed to include in the complaint an element of the charged offense. 163 P.3d at 1138. Consequently, although the charge was labeled a class 4 felony, it was substantively a class 5 felony. Id. at 1138-39. On the first day of trial, the prosecution cleared up the discrepancy by proffering an elemental jury instruction for a class 5 felony. Id. at 1139. The parties then proceeded-both in terms of the presentation of evidence and their trial strategy-consistent with that tendered instruction, and the court ultimately gave that instruction to the jury. Id.
¶30 After the jury returned a guilty verdict pursuant to the class-5-felony elemental instruction, the sentencing judge (a different judge than the one who presided over the trial) sentenced Medina for the unnoticed and unproven class 4 felony. Id. As a result, Medina was convicted of and sentenced for a crime that wasn't even charged, let alone proved (either to the jury or to a judge). Id. Under these circumstances, we determined that the trial court "entered its own conviction and sentence . . . instead of determining the punishment warranted by the jury's guilty verdict." Id. at 1140. We held that the trial court violated Medina's due process rights because it "essentially judged Medina guilty of a new and different crime" without so much as providing him adequate notice of it. Id. at 1141. And, deeming the error structural, we automatically reversed the judgment of conviction and sentence and remanded for resentencing on the class 5 felony (the crime actually charged and proved). Id. at 1141-42.
¶31 A Linnebur error is different. See Viburg II, ¶ 27, 500 P.3d at 1130. In a Linnebur error, the prosecution properly charges the defendant with felony DUI by including all the elements of the offense, and the defendant is fully on notice that he is charged with and could be convicted of felony DUI. Id. Further, evidence of every element of felony DUI is presented and there is a finding that each element has been proved by the prosecution, even if one of those elements is incorrectly proved only by a preponderance of the evidence to the judge. See id. But despite the error in failing to submit the element of the defendant's relevant convictions to the jury, the trial court neither adjudicates the defendant guilty of an unnoticed and unproven crime nor enters a sentence for such a crime. Id.
We concluded in Viburg II that, following the automatic reversal of Viburg's conviction and sentence for felony DUI (as required by Linnebur), it did not violate due process principles or the Double Jeopardy Clause to allow the prosecution to retry him on remand for felony DUI. Viburg II, ¶ 27, 500 P.3d at 1130.
¶32 In our view, a Linnebur error is more akin to the error we reviewed in Griego v. People, 19 P.3d 1, 2 (Colo 2001), where the trial court failed to instruct the jury on an element of the charged offense See People v Carter, 2021 COA 29, ¶¶ 65-67, 486 P.3d 473, 486 (J Jones, J, concurring dubitante). Guided by Neder v. United States, 527 U.S. 1, 8 (1999) (concluding that "a jury instruction that omits an element of the offense" is an error that "differs markedly from the constitutional violations we have found to defy harmless-error review"), we held in Griego that "when a trial court misinstructs the jury on an element of an offense, either by omitting or misdescribing that element, that error is subject to constitutional harmless or plain error analysis and is not reviewable under structural error standards." Griego, 19 P.3d at 8.
¶33 We reaffirmed this principle a decade later in Tumentsereg v. People, 247 P.3d 1015, 1018 (Colo. 2011). There, we characterized the rule in Griego as "well-settled." See id. ("[E]rror in the form of a misdescription or omission of an element of an offense does not, for that reason alone, constitute structural error.").
¶34 Notably, Medina explicitly drew the distinction we highlight today between the error in that case (where the defendant was convicted and sentenced for a crime that was unnoticed and unproven) and the errors in Neder and Griego (where the defendants were convicted and sentenced following the omission of an element from a jury instruction). Medina, 163 P.3d at 1137-38; see also Carter, ¶ 67, 486 P.3d at 486 (J. Jones, J., concurring dubitante). Distinguishing Neder and Griego, we observed that the errors in those cases were trial errors requiring application of the appropriate outcome-determinative standard of reversal, while the error in Medina's trial, by contrast, was structural and required automatic reversal. Medina, 163 P.3d at 1137-38; see also Carter, ¶ 67, 486 P.3d at 486 (J. Jones, J., concurring dubitante).
Tumentsereg postdated Medina.
¶35 In short, Medina doesn't support the proposition for which we cited it in Linnebur-that the trial court's error in failing to present the element of Linnebur's relevant convictions to the jury was structural and subject to automatic reversal. And Neder, Griego, and Tumentsereg, which we overlooked in Linnebur, persuade us to conclude now that the error before us in Linnebur was a trial error subject to the appropriate outcome-determinative standard.
¶36 Of course, "stare decisis," a Latin term meaning "to stand by things decided," generally forces us to be faithful to our prior decisions. People v. Kembel, 2023 CO 5, ¶ 43, 524 P.3d 18, 27 (quoting Stare Decisis, Black's Law Dictionary (11th ed. 2019)). There are compelling reasons for this doctrine: "Stare decisis . . . promotes the evenhanded, predictable, and consistent development of legal principles, fosters reliance on judicial decisions, and contributes to the actual and perceived integrity of the judicial process." Love v. Klosky, 2018 CO 20, ¶ 14, 413 P.3d 1267, 1270 (quoting Payne v. Tennessee, 501 U.S. 808, 827 (1991)).
¶37 But we conclude that this is one of those rare cases in which "'the iron grip of stare decisis' should be loosened." Kembel, ¶ 44, 524 P.3d at 27 (quoting United States v. Reveron Martinez, 836 F.2d 684, 687 n.2 (1st Cir. 1988)). "[W]e are faithful to a preexisting rule of law unless we are clearly convinced that the rule was originally erroneous or is no longer sound because of changing conditions and that more good than harm will come from departing from [that] precedent." Id. at ¶ 43, 524 P.3d at 27 (quoting McShane v. Stirling Ranch Prop. Owners Ass'n, 2017 CO 38, ¶ 26, 393 P.3d 978, 984).
¶38 It is clear to us now that we erred in Linnebur. It happens. Appellate court judges are no more infallible than our esteemed colleagues on the trial court bench. See In re Bushkin Assocs., Inc., 864 F.2d 241, 245 (1st Cir. 1989) ("To err is human, and judges-trial and appellate alike-are not infallible."). As Rory Charles Graham reminds us in his hit song, we're "only human after all." Rag'n'Bone Man, Human, on Human (Columbia Records 2017).
¶39 This case gives us an opportunity for course correction. And we are firmly convinced that more good than harm will come from reversing direction. Linnebur strayed from Neder, Griego, and Tumentsereg. And it did so without discussing stare decisis. Today we simply right the ship. See Rios-Vargas v. People, 2023 CO 35, ¶ 43, 532 P.3d 1206, 1216.
¶40 Having determined that the error in this case was not structural, the question that naturally follows is: What is the applicable standard of reversal? Asked differently, how do we decide whether the error that occurred in Crabtree's trial requires reversal? Because Crabtree did not object and the error was thus unpreserved, we must review for plain error. See Hagos, ¶ 14, 288 P.3d at 120. We thus segue to the second issue we agreed to review: Whether the division erred in the way it treated the plainness prong of plain error review under Crim. P. 52(b).
B. Did the Division Incorrectly Apply a Time-of-Appeal Rule in Considering Plainness?
¶41 Our appellate courts normally won't correct a trial error if the defendant failed to call the court's attention to it by uttering an objection. But Crim. P. 52(b) contains an exception to this rule. It states that "[p]lain errors . . . affecting substantial rights may be noticed [on appeal] although they were not brought to the attention of the [trial] court." Crim. P. 52(b). Thus, Crim. P. 52(b) permits review if (1) there is an error, (2) that is plain, and (3) that affects the defendant's substantial rights.
¶42 An error is "plain" within the meaning of Crim. P. 52(b) if it is "obvious." People v. Miller, 113 P.3d 743, 750 (Colo. 2005). More specifically, the plainness prong demands that the error be "'so clear-cut, so obvious,' a trial judge should be able to avoid it without benefit of objection." Romero v. People, 2017 CO 37, ¶ 6, 393 P.3d 973, 976 (quoting People v. Ujaama, 2012 COA 36, ¶ 42, 302 P.3d 296, 304). Consequently, to be deemed plain, an error must contravene a clear statutory command, a well-settled legal principle, or established Colorado case law. Scott, ¶ 16, 390 P.3d at 835. Conversely, when Colorado statutory law or case law would not have alerted the trial judge to an unobjected-to error, the error cannot be deemed plain. Id. at ¶ 17, 390 P.3d at 835.
¶43 But Crim. P. 52(b) requires more than plainness. The unpreserved error must also have affected "the substantial rights of the accused." People v. Stewart, 55 P.3d 107, 120 (Colo. 2002). We have recognized that a plain error requires reversal if an appellate court, after reviewing the trial record in its entirety, can say with fair assurance that the error "so undermined the fundamental fairness of the trial itself as to cast serious doubt on the reliability of the judgment of conviction." Wilson v. People, 743 P.2d 415, 420 (Colo. 1987). In so doing, we have explained that "the plain-error rule of Crim. P. 52(b) is calculated to temper the contemporaneous-objection requirement in the interests of permitting an appellate court to correct particularly egregious errors." Id.
In many recent cases, rather than declare that Crim. P. 52(b) applies if (1) there is an error, (2) that is plain, and (3) that affects the defendant's substantial rights, and that a plain error is reversible if it so undermines the fundamental fairness of the trial itself as to cast serious doubt on the reliability of the judgment of conviction, see, e.g., Hagos, ¶ 18, 288 P.3d at 120; Miller, 113 P.3d at 750, we have treated the statement about the fundamental fairness of the trial as an additional prong of the plain error standard, see, e.g., Scott, ¶ 15, 390 P.3d at 835; Romero, ¶ 6, 393 P.3d at 976. We perceive no substantive difference between these two iterations; rather, we see any distinction as akin to the old "po-tay-to/po-tah-to," "to-may-to/to-mah-to" debate. Fred Astaire & Ginger Rogers, Let's Call the Whole Thing Off, in Shall We Dance? (RKO Radio Pictures, Inc. 1937) ("You like po-tay-to and I like po-tah-to; You like to-may-to and I like to-mah-to.").
¶44 Here, we are concerned only with the plainness prong. We limit our analysis accordingly.
¶45 The division acknowledged that the trial court's error in omitting from the jury's consideration the element of Crabtree's relevant convictions was not obvious at the time of his trial. Crabtree, ¶ 41, 519 P.3d at 423. Indeed, it couldn't have been. Far from contravening a clear statutory command, a well-settled legal principle, or established Colorado case law, the trial court followed the only relevant authority in effect at the time-the court of appeals' decision in Gwinn. Because the trial court was bound by Gwinn, it had no choice but to adhere to that case's holding. As we noted in Scott, trial judges are "bound to follow the decisions of the appellate courts and cannot generally be faulted for not departing from that authority sua sponte." ¶ 17, 390 P.3d at 835. We certainly can't expect trial judges to be clairvoyant and predict that the law they are bound by will change.
¶46 Still, the division concluded that the error was plain because it was obvious at the time of Crabtree's appeal. By that time, our decision in Linnebur had entered the picture.
¶47 In taking a time-of-appeal approach, the division relied on the Supreme Court's opinions in Johnson and Henderson. Crabtree, ¶¶ 48-50, 519 P.3d at 424-25. The division reasoned that three other divisions had already adhered to Johnson, and Henderson was simply an extension of Johnson. Id. at ¶¶ 44-47, 519 P.3d at 424. Recall that in Johnson, the Court held that "where the law at the time of trial was settled and clearly contrary to the law at the time of appeal[,] it is enough that an error be 'plain' at the time of appellate consideration," 520 U.S. at 468, and in Henderson, the Court went a step further and held that, so long as an error is obvious at the time of appeal, it doesn't matter whether the law was settled or unsettled at the time of trial, 568 U.S. at 273-75.
¶48 But our court has never embraced Johnson, and we're not bound by the three court of appeals' decisions that have followed it. Thus, we disagree with the division's statement that, following those three cases, "the rule [is] clear: Johnson governs plain error review in Colorado when the applicable law was settled at the time of trial but not when the applicable law was not settled at the time of trial." Crabtree, ¶ 47, 519 P.3d at 424.
Those three opinions and any like-minded ones from the court of appeals are overruled to the extent they are inconsistent with this opinion.
We pass no judgment on the division's conclusion that the relevant law in Colorado was settled at the time of Crabtree's trial based on the decision in Gwinn.
¶49 We decline to embrace Johnson and its progeny today. Despite the significant similarities between Crim. P. 52(b) and Fed. R. Crim. P. 52(b), we have historically employed a different plain error standard than the one the Supreme Court uses. In contrast to Colorado's Crim. P. 52(b) standard, the Supreme Court's Fed. R. Crim. P. 52(b) standard authorizes (but does not require) appellate courts to grant a defendant relief if (1) there is an error; (2) that is plain (i.e., that is clear or obvious); (3) that affects the defendant's substantial rights; and (4) that seriously affects the fairness, integrity, or public reputation of judicial proceedings. See United States v. Olano, 507 U.S. 725, 732-36 (1993).
¶50 There are two significant differences between the two plain error standards. First, unlike the federal standard, Colorado's standard involves mandatory, not discretionary, relief. That is, if a Colorado appellate court concludes that an unpreserved error was obvious or clear-cut and that it affected the defendant's substantial rights, it must grant relief under Crim. P. 52(b). Second, the fourth prong of the federal standard (Olano's fourth prong) is conspicuously absent from Colorado's-unlike federal courts, we don't consider whether an error seriously affects the fairness, integrity, or public reputation of judicial proceedings. In Olano, the Supreme Court characterized this prong as the one "that should guide the exercise of remedial discretion"-that is, the benchmark on which appellate courts should rely to determine when "a plain forfeited error affecting substantial rights" should be corrected. Id. at 736.
¶51 Given the different analytical frameworks in Colorado's plain error cases, on the one hand, and the Supreme Court's plain error cases, on the other, we are not bound by Johnson and Henderson. Johnson and Henderson are simply inapposite.
¶52 Having said that, we acknowledge that we've agreed on multiple occasions to consider whether we should get behind the time-of-appeal rule ushered in by Johnson and refined by Henderson, though we have ultimately declined to address it each time after concluding it was unnecessary to do so. See Thompson v. People, 2020 CO 72, ¶ 53, 471 P.3d 1045, 1057; Campbell, ¶ 38, 464 P.3d at 767; Howard-Walker v. People, 2019 CO 69, ¶ 49, 443 P.3d 1007, 1016; Garcia v. People, 2019 CO 64, ¶ 28, 445 P.3d 1065, 1070; Romero, ¶ 1 n.1, 393 P.3d at 975 n.1. But, contrary to the division's suggestion, that doesn't mean that Colorado lacks a rule on the temporal scope of plain error review pursuant to Crim. P. 52(b). To be sure, we don't write on a tabula rasa today.
¶53 Historically, whenever we've reviewed for plain error, we've focused on the plainness of the error at the time of trial, not at the time of appeal. Indeed, the parties do not point us to any case in which our plain error review was premised on the plainness of the error at the time of review-and we're aware of none. Further, dating back to Scott, we've explained that plainness refers to how obvious or clear-cut an error is at the time it is made. ¶ 16, 390 P.3d at 835.
¶54 Scott is not a lone wolf; it's a shepherd with a flock-we count over a dozen sheep in its pasture, including some that hail from the court of appeals. See, e.g., People v. Seymour, 2023 CO 53, ¶ 40 n.9, 536 P.3d 1260, 1274 n.9; Campbell, ¶ 25, 464 P.3d at 765; People v. Delgado, 2019 CO 82, ¶ 32, 450 P.3d 703, 708; Cardman v. People, 2019 CO 73, ¶ 34, 445 P.3d 1071, 1082; People v. Koper, 2018 COA 137, ¶ 43, 488 P.3d 409, 417; People v. Wambolt, 2018 COA 88, ¶ 70, 431 P.3d 681, 695. Thus, the question we've left open is whether to change from the time-of-trial rule to the time-of-appeal rule, not (as the division surmised) whether to adopt the time-of-appeal rule to bridge a gap in our analytical framework for plain error review.
¶55 Like other defendants have done in the past, Crabtree urges us to replace our time-of-trial rule with Henderson's time-of-appeal rule. But this time, the question is properly teed up and we have occasion to answer it.
¶56 So, should we replace our time-of-trial rule with Henderson's time-of-appeal rule? The text of Crim. P. 52(b) offers little assistance. It simply provides that an appellate court may consider "[p]lain errors," language that leaves the temporal question open. See Henderson, 568 U.S. at 272 (making a similar observation with respect to the text of Fed. R. Crim. P. 52(b)).
¶57 For several reasons unrelated to the text of the rule, we turn down Crabtree's invitation to veer away from our well-trodden sheep track. Instead, we remain faithful to our case law.
¶58 First, the Supreme Court in Henderson endorsed a time-of-appeal approach in part because of the unfairness of treating defendants in different federal circuits unevenly based on the state of the law in each circuit at the time an error occurs. Id. at 274-75. In Colorado, however, this concern is irrelevant because the law is, of course, uniform across the state. In general, all defendants tried in Colorado are affected by the same law at the same time, regardless of the judicial district in which a trial occurs.
¶59 Second, in Henderson, the Court observed that a time-of-trial rule "creates unfair and anomalous results." Id. at 276-77. The Court was mindful that the basic purpose of Fed. R. Crim. P. 52(b) is to establish "a fairness-based exception to the general requirement that an objection be made at trial." Id. at 276. In the Court's view, "a rigid and undeviating judicially declared practice under which courts of review would invariably and under all circumstances decline to consider all questions which had not previously been specifically urged would be out of harmony with . . . the rules of fundamental justice." Id. at 272 (alteration in original) (quoting Olano, 507 U.S. at 732). But in Colorado, Crim. P. 35(c)(1) alters the equation by mitigating any possible unfairness.
¶60 Crim. P. 35(c)(1) provides a postconviction mechanism to address a "significant change in the law" during the pendency of a direct appeal. See Crim. P. 35(c)(1) ("If, prior to filing for relief pursuant to this paragraph (1), a person has sought appeal of a conviction . . . that person may file an application for postconviction review upon the ground that there has been a significant change in the law . . . allowing in the interests of justice retroactive application of the changed legal standard."). There is no federal analogue to Crim. P. 35(c)(1). Thus, but for a time-of-appeal rule under Fed. R. Crim. P. 52(b), defendants may be out of luck in federal court when there is a significant change in the law that favors them while their cases are pending on appeal. Not so in Colorado state courts.
¶61 And contrary to Crabtree's contention, Colorado courts have interpreted Crim. P. 35(c)(1) to apply to significant changes in the law outside of the context of amendatory legislation. See People v. Allen, 843 P.2d 97, 100-01 (Colo.App. 1992), rev'd on other grounds, 868 P.2d 379 (Colo. 1994) (applying the rule based on changes in the law deriving from a judicial decision). Consequently, the scope of the rule is not nearly as narrow as Crabtree posits.
¶62 Crabtree is correct, though, that relief under Crim. P. 35(c)(1) is in the discretion of trial courts when it is "in the interests of justice." But that's hardly cause for concern-we routinely entrust trial courts with weighty discretionary decisions, and time and again they prove that they are up to the task.
¶63 Lastly, in dispelling the "'plain error' floodgates" concern that accompanies a time-of-appeal rule, the Supreme Court in Henderson relied in part on the discretion vested in federal appellate courts to deny relief unless the error "would, in fact, seriously affect the fairness, integrity, or public reputation of judicial proceedings." 568 U.S. at 277-78. In Henderson, the Government asserted that a time-of-appeal rule would lead to many more claims of plain error because appellate courts regularly clarify the law through their opinions. Id. at 278. According to the Government, a time-of-appeal standard would invite defendants who never objected at trial to insist that the appellate courts evaluate their cases according to the new rule. Id. And that, in turn, hypothesized the Government, would water down "plain error" into "simple 'error.'" Id.
¶64 The Supreme Court was unconvinced for several reasons, including the presence of "other screening criteria." Id. Specifically, the Court reasoned that, under the federal plain error standard, the error must have affected the defendant's substantial rights and must have seriously affected the fairness, integrity, or public reputation of judicial proceedings. Id. Using these two criteria, noted the court, may well mean that a defendant's failure to object at trial despite unsettled law will "count against" the grant of relief pursuant to plain error review. Id. at 278-79.
¶65 Colorado's plain error standard includes one of these two criteria (whether the error affects the defendant's substantial rights) but not the other (Olano's fourth prong-whether the error seriously affects the fairness, integrity, or public reputation of judicial proceedings). And the missing criterion is the one federal courts use to exercise their discretion in deciding whether to grant relief when an error is plain and affects the defendant's substantial rights. This is a big deal.
Although a couple of our opinions have seemingly incorporated Olano's fourth prong, see Hagos, ¶ 18, 288 P.3d at 120 (stating that plain errors must "seriously affect the fairness, integrity or public reputation of judicial proceedings" (quoting United States v. Young, 470 U.S. 1, 15 (1985)); Domingo-Gomez v. People, 125 P.3d 1043, 1053 (Colo. 2005) (stating that, under the plain error standard, "[a] reviewing appellate court must inquire into whether the errors seriously affected the fairness or integrity of the trial"), we have never formally adopted it as part of our analytical framework.
¶66 Just a few years ago, the Connecticut Supreme Court declined to adopt Henderson's time-of-appeal rule precisely because Connecticut's analytical framework for plain error review differs from the one employed by the federal courts. See State v. Turner, 224 A.3d 129, 145 (Conn. 2020). Connecticut's plain error methodology may not mirror Colorado's, but Turner is nevertheless instructive:
This court has declined to adopt the federal plain error rule, . . . concluding that federal case law is "inapposite and unpersuasive" in determining the scope of plain error review. This is because of the "fundamental differences" between federal and state law regarding the plain error doctrine. "Under federal law" . . . clear error is just one aspect of the . . . plain error doctrine, even if measured as of the time of the appeal. "By contrast . . . Connecticut's plain error doctrine is a rule of reversibility, mandating reversal when plain error is found." Unlike federal courts, Connecticut appellate courts do not have discretion to reverse a conviction for plain error, and the defendant does not ask this court to grant appellate courts this discretion.
In light of this distinction between the federal plain error doctrine and Connecticut's plain error doctrine, we . . . decline to extend our plain error doctrine to errors that were not clear at the time of trial and require reversal in cases in which both the trial court and the parties properly applied the law existing at the time of trial.Id. (fourth alteration in original) (emphases added) (citations omitted) (quoting State v. McClain, 155 A.3d 209, 216 n.8 (Conn. 2017)).
¶67 Like the Connecticut Supreme Court, we are uncomfortable with the prospect of adopting Henderson's time-of-appeal rule given the crucial differences between the federal plain error standard and ours. These differences illustrate the danger in selectively borrowing a single aspect of the federal plain error standard while disregarding the rest. Inasmuch as our time-of-trial rule was not originally erroneous and has not ceased to be sound, we are not at liberty to divert from principles of stare decisis and shift to a time-of-appeal rule.
We are not persuaded otherwise by Crabtree's speculation that continuing on the time-of-trial path will inundate trial courts with a deluge of objections that are meritless under existing law. As mentioned earlier, in situations where an objection would be futile under then-existing precedent, a defendant in Colorado state court may seek relief pursuant to Crim. P. 35(c)(1) should a significant change in the law occur during the pendency of the direct appeal. This rule may well explain why our trial courts haven't experienced a downpour of objections over the many years during which our time-of-trial standard has been applied by our court and most of the divisions of the court of appeals.
¶68 Why not embrace the Supreme Court's plain error analytical framework in its entirety then? After all, we have endeavored to interpret our state rules in conformity with federal courts' interpretations of similar rules. See Warne v. Hall, 2016 CO 50, ¶ 12, 373 P.3d 588, 592 (dealing with the rules of civil procedure); Crumb v. People, 230 P.3d 726, 731 n.5 (Colo. 2010) (dealing with the rules of criminal procedure); People v. Melendez, 102 P.3d 315, 319 (Colo. 2004) (dealing with the rules of evidence); see also Novotny, ¶ 21, 320 P.3d at 1201 (stating that we have now largely accepted the structural error/trial error dichotomy adopted by the Supreme Court). And consistent with that practice, members of this court have been receptive to an alignment with the Supreme Court's plain error framework. See Maestas v. People, 2019 CO 45, ¶ 35, 442 P.3d 394, 401 (Samour, J., concurring in the judgment only).
¶69 We carefully considered this option. But for various reasons, we opt to stick with the status quo.
¶70 Neither party in this case is asking us to discard our plain error standard for the federal one. Indeed, there doesn't seem to be much appetite for the federal standard in the defense bar. In each of the cases in which the defense has asked us to adopt Henderson's time-of-appeal approach, it has opposed any efforts to revamp our plain error standard in its entirety. Nor are the People banging the drum to supplant our plain error standard with the federal one. Instead, they merely urge, in the alternative, that if we are inclined to become Henderson disciples, we drop our standard and get on board with the federal one. Given today's decision, that request is obsolete.
¶71 Perhaps most importantly, our plain error analytical framework has proven effective and efficient over the course of many years, and neither party has identified deficiencies warranting a complete renovation. Such an overhaul would undoubtedly be greatly disruptive. And for what? Crim. P. 35(c)(1) already guards against the injustices that Henderson's time-of-appeal rule seeks to avert.So, because it ain't broke, we abstain from fixing it.
We express no opinion as to whether a defendant seeking a limited remand under Crim. P. 35(c)(1) must join any Crim. P. 35(c)(2) claims or whether a defendant's failure to do so bars as successive any Crim. P. 35(c)(2) claims subsequently brought. Those questions are not before us in this case.
C. Did the Division Incorrectly Determine that the Error Here Amounted to Plain Error Under Crim. P. 52(b)?
¶72 The parties agree that the unobjected-to error by the trial court-omitting the element of Crabtree's relevant convictions from the jury's consideration-was not plain at the time it was made. And because we conclude that relief under Crim. P. 52(b) is only available if the error is plain at the time it is made, Crabtree cannot prevail.
IV. Conclusion
¶73 For the foregoing reasons, we reverse the division's judgment. We remand with instructions to return the case to the trial court for reinstatement of Crabtree's felony DUI conviction and sentence.
JUSTICE HOOD concurred in part and dissented in part. JUSTICE GABRIEL dissented.
JUSTICE HOOD, concurring in part and dissenting in part.
¶74 I write separately to join Part III.A of the majority opinion and Part II.B of Justice Gabriel's dissenting opinion. In summary, I agree with the majority that we should review what occurred here for plain error (even if it means overruling our recent precedent suggesting otherwise), but I also agree with Justice Gabriel that plain-error review should be premised on the case law in effect at the time of appeal. In the interest of avoiding redundancy, I embrace my colleagues' thoughtful rationales in the referenced portions of their opinions for my hybrid conclusion. Like Justice Gabriel, I would affirm the judgment of the court of appeals. Therefore, I respectfully concur in part and dissent in part.
JUSTICE GABRIEL, dissenting.
¶75 Today, the majority concludes that it is not structural error for a trial court to sentence a defendant for felony driving under the influence ("DUI"), notwithstanding the fact that the jury did not convict the defendant of felony DUI. Maj. op. ¶¶ 1, 37-40. In so concluding, the majority overrules what we held just four years ago in Linnebur v. People, 2020 CO 79M, ¶¶ 32-33, 476 P.3d 734, 741-42, and disregards two decades of this court's precedent.
¶76 The majority then compounds what I perceive to be its foregoing error by refusing to apply the law in effect at the time a case is before us on direct (as compared to postconviction) review, notwithstanding (1) long-settled precedent that on direct review, appellate courts apply the law in effect at the time of appeal; and (2) United States Supreme Court case law that is directly on point.
¶77 Because I believe that both conclusions are incorrect and produce unjust results, I would affirm the judgment of the division below, although on somewhat different grounds.
¶78 Accordingly, I respectfully dissent.
I. Factual Background
¶79 The facts that are pertinent to my analysis are undisputed.
¶80 In June 2017, the People charged Charles James Crabtree with (1) misdemeanor DUI and (2) felony DUI, and they alleged that this was at least his fourth alcohol-related offense. The case proceeded to trial, at which the trial court instructed the jury on only the elements of misdemeanor DUI. The jury ultimately found Crabtree guilty of that offense, after which the court entered the verdict and dismissed the jury.
¶81 The next day, the court held a hearing, and it found by a preponderance of the evidence that Crabtree had committed four prior alcohol-related offenses. Based on that finding, the court elevated Crabtree's conviction from a misdemeanor to a class four felony. Crabtree did not object to this procedure because, at the time, prior convictions were deemed to be sentence enhancers that were to be determined by the court, rather than elements of the offense to be found by the jury. See People v. Gwinn, 2018 COA 130, ¶ 39, 428 P.3d 727, 736. In accordance with its findings, the trial court sentenced Crabtree to both misdemeanor and felony DUI, and Crabtree appealed.
¶82 While Crabtree's appeal was pending, we decided Linnebur, ¶ 31, 476 P.3d at 741, in which we concluded that for a felony DUI, prior convictions are elements of the crime that must be proved to the jury beyond a reasonable doubt, rather than sentence enhancers that the court may find by a preponderance of evidence.
¶83 In light of our intervening decision in Linnebur, Crabtree then argued on appeal, for the first time, that his felony DUI conviction must be vacated because the prosecution had not proved his prior convictions to the jury beyond a reasonable doubt. In a unanimous, published decision, a division of our court of appeals reversed the felony DUI conviction and remanded the case to the trial court, concluding that that court had plainly erred by not submitting to the jury the question of whether Crabtree had prior alcohol-related convictions and by not requiring the People to prove Crabtree's prior convictions beyond a reasonable doubt. People v. Crabtree, 2022 COA 73, ¶ 1, 519 P.3d 415, 417.
¶84 In so ruling, the division rejected the People's contention that appellate courts must evaluate the obviousness of the error as of the time of trial, rather than as of the time of appeal. Id. at ¶ 51, 519 P.3d at 425. The division observed that the Supreme Court had made clear in Johnson v. United States, 520 U.S. 461, 468 (1997), that when, as here, "the law at the time of trial was settled and clearly contrary to the law at the time of appeal[,] it is enough that an error be 'plain' at the time of appellate consideration." Crabtree, ¶ 43, 519 P.3d at 424 (quoting Johnson, 520 U.S. at 468). The division further determined that Crabtree had satisfied both prongs of plain error review (i.e., obviousness and substantiality) because (1) in light of Linnebur, the procedure employed in Crabtree's trial was clearly contrary to the law at the time of appeal; and (2) Crabtree was entitled to benefit on direct appeal from that change in the law. Id. at ¶ 50, 519 P.3d at 425.
II. Analysis
¶85 I begin by addressing why I believe that it is structural error for a trial court to convict a defendant of felony DUI when the jury was instructed on, and returned a verdict for, misdemeanor DUI. I then proceed to explain why I believe that an appellate court, on direct review, should assess plain error by applying the law in effect at the time of appeal, rather than the law in effect at the time of trial.
A. Structural Error
¶86 Our precedent recognizes two types of constitutional errors: trial errors and structural errors. People v. Vigil, 127 P.3d 916, 929 (Colo. 2006). Trial errors are those that occur in the trial process itself, and an appellate court reviews such errors for either harmless or plain error, depending on whether the issue was preserved. Id. Structural errors, in contrast, are "defects affecting the framework within which the trial proceeds." Id. (quoting People v. Miller, 113 P.3d 743, 749 (Colo. 2005)). Because structural errors infect an entire trial and necessarily render the trial fundamentally unfair, Neder v. United States, 527 U.S. 1, 8 (1999), they cannot be reviewed under a harmless or plain error standard; rather, they require automatic reversal, Miller, 113 P.3d at 749.
¶87 A defendant's right to a trial by jury is a "structural guarantee" because it "reflects a fundamental decision about the exercise of official power" and a concern about entrusting plenary powers over a defendant's life and liberty to only one judge or to a group of judges. Griego v. People, 19 P.3d 1, 8 (Colo. 2001) (quoting People v. Vance, 933 P.2d 576, 580 (Colo. 1997), disapproved on other grounds by Griego, 19 P.3d at 8). If a court sentences a defendant to "a crime different from that on which a jury's guilty verdict is based," then the court has essentially reconsidered the defendant's guilt following the verdict. Medina v. People, 163 P.3d 1136, 1138 (Colo. 2007). In doing so, the court has committed a structural error, and the sentence must be vacated. Id.
¶88 In Medina, 163 P.3d at 1138, we granted certiorari to decide whether the plain error standard of review applies when a trial court instructs the jury on the elements of a lesser offense than the one for which the defendant was ultimately sentenced. In that case, the trial court instructed the jury on the elements of class five felony accessory, and the jury ultimately returned a guilty verdict. Id. at 1139. After entering the verdict, however, the trial court concluded that Medina had been convicted of the greater offense of class four accessory, and it sentenced her accordingly. Id. Medina appealed, and a division of our court of appeals affirmed her conviction, concluding that although the jury instruction contained a "misdescription" of an element of the class four accessory charge, Medina did not object to the instruction at trial and, thus, plain error review applied. Id. at 1139-40. The division ultimately upheld the conviction because it determined that the misdescribed element did not undermine the sentence's validity and, thus, did not rise to the level of plain error. Id. at 1140.
¶89 We granted certiorari and vacated the class four accessory conviction, concluding that structural, and not plain, error analysis applies "when a sentencing court enters a conviction different from that provided for by a jury's finding of guilt based upon a jury instruction that correctly and completely describes all the elements of a less serious offense rather than misdescribing or omitting an element of an offense." Id. at 1138, 1142 (emphasis added). In so concluding, we reasoned that when a trial court sentences a defendant to a crime different from that authorized by the jury's verdict, the court essentially adjudges the defendant guilty of "a new and different crime." Id. at 1141. Because such a determination violates a defendant's due process and Sixth Amendment rights to a trial by jury, the entry of such a conviction is "an error that infect[s] each and every aspect" of the sentence and constitutes structural error requiring that the sentence be vacated. Id. at 1138, 1141.
¶90 We subsequently followed Medina in Linnebur, ¶¶ 1-2, 32, 476 P.3d at 735, 741, a case in which we were tasked with deciding whether the fact of prior convictions was an element of felony DUI that must be proved to the jury beyond a reasonable doubt, rather than a sentence enhancer that the trial judge could find by a preponderance of the evidence. In that case, the jury had been instructed on, and had found Linnebur guilty of, driving while ability impaired ("DWAI") and DUI per se. Id. at ¶¶ 4-5, 476 P.3d at 735. After entering the verdict and dismissing the jury, the court held a hearing to determine whether Linnebur had at least three prior alcohol-related convictions that would support felony convictions. Id. at ¶ 5, 476 P.3d at 735-36. The court so found and convicted Linnebur of felony DWAI, after which it merged the DUI per se conviction into the felony conviction. Id. at ¶ 5, 476 P.3d at 736. (I note that although Linnebur involved a felony DWAI conviction, we noted in that case that the DWAI statute includes language identical to that included in the DUI statute elevating a DWAI conviction to a felony offense after three prior convictions, id. at ¶ 5 n.1, 476 P.3d at 735 n.1, and we sometimes referred to Linnebur's conviction in that case as a conviction for felony DUI; for convenience and to avoid confusion, I will do the same here.) Linnebur appealed, and a division of our court of appeals affirmed, reasoning that the fact of prior convictions is a sentence enhancer, not an element, of felony DUI and, thus, the convictions could properly be determined by a court by a preponderance of the evidence instead of by a jury beyond a reasonable doubt. Id. at ¶ 6, 476 P.3d at 736.
¶91 We granted certiorari and reversed the felony conviction, concluding that the fact of prior alcohol-related convictions was a substantive element of the felony offense that must be proved to a jury beyond a reasonable doubt. Id. at ¶ 31, 476 P.3d at 741. Specifically, citing Medina, 163 P.3d at 1141-42, we determined that Linnebur's felony conviction could not stand because he had been "sentenced for a crime different from the one on which the jury's verdict was based." Id. at ¶ 32, 476 P.3d at 741. In reaching this conclusion, we reasoned that allowing a defendant to be sentenced to a felony based on a fact that was found by the court by a preponderance of the evidence, despite the defendant's having been tried to the jury only for a misdemeanor, would be so unfair that it would risk "running afoul of the Sixth Amendment." Id. at ¶ 29, 476 P.3d at 741.
¶92 I believe that we got it right in Linnebur, and I perceive no basis for us to depart from the rule that we adopted so recently, without any persuasive reason for doing so. Our precedent is clear that when a judge rather than the jury makes a finding on which a conviction is based, this is the very essence of structural error-an error that goes to the very heart of the trial's structure. Here, because the judge, and not the jury, found that Crabtree had committed at least three prior alcohol-related offenses, and the trial court used those prior convictions as the basis for elevating Crabtree's conviction from a misdemeanor to a felony, the court essentially reconsidered Crabtree's guilt after the jury verdict had been entered. This error affected the framework of the trial by removing the determination of guilt from the jury and placing it in the hands of a single judge. Doing so necessarily rendered the trial fundamentally unfair, and, thus, in my view, Crabtree's felony DUI conviction must be vacated.
¶93 I am not persuaded otherwise by the majority's conclusion that the error here was akin to the instructional errors in Neder, 527 U.S. at 9, and Griego, 19 P.3d at 4. See Maj. op. ¶¶ 32-35.
¶94 In Neder, 527 U.S. at 6-7, Neder appealed his convictions for filing false income tax returns and for federal mail, wire, and bank fraud, alleging that the trial court had committed structural error by instructing the jury that it need not consider the issue of the materiality of any false statements because materiality was not a question for the jury to decide. Neder contended that because materiality was an element of the offenses for which he was charged, his convictions could not stand, given that the jury had not found all of the elements beyond a reasonable doubt. Id. at 11. Although the Supreme Court agreed that the trial court had erred in deciding the materiality element itself rather than submitting that question to the jury, it concluded that a jury instruction that omits an element of the offense does not always render a trial fundamentally unfair and, thus, such an omission should be reviewed for harmless error. Id. at 8-9. The Court then concluded that the instructional omissions were harmless because the omitted elements were supported by uncontroverted evidence at trial. Id. at 18-20.
¶95 Similarly, in Griego, 19 P.3d at 8, this court held that "when a trial court misinstructs the jury on an element of an offense, either by omitting or misdescribing that element, that error is subject to constitutional harmless or plain error analysis and is not reviewable under structural error standards." (Emphases added.) In that case, Griego was charged with driving after revocation. Id. at 2. He conceded that he had driven while his license was under revocation, but he claimed that he did not have knowledge of the revocation. Id. At trial, Griego requested that the trial court instruct the jury on the mental state of "knowingly." Id. at 3. The court, however, refused to provide such an instruction, reasoning that because the stock elemental instruction listed the element as "knowledge," providing the definition of "knowingly" would not be helpful to the jury. Id. The jury convicted Griego of driving after revocation, a division of our court of appeals affirmed, and we granted certiorari. Id. at 4. We ultimately affirmed Griego's conviction, concluding that although the trial court had constitutionally erred in failing to instruct the jury on the mental state of "knowingly," this error was harmless beyond a reasonable doubt. Id.
¶96 This case is unlike either Neder or Griego. In those cases, the trial courts had failed to instruct the jury on or misdescribed an element of the offense for which the defendant was charged and ultimately convicted. Here, in contrast, the trial court never instructed the jury on the offense for which Crabtree was ultimately sentenced. Failing to instruct on an offense itself is not the same as omitting or misdescribing an element of a charged offense.
¶97 Nor am I persuaded that People v. Viburg, 2021 CO 81M, 500 P.3d 1123 ("Viburg II"), on which the majority also relies, Maj. op. ¶¶ 28-31, somehow signaled a retreat from Medina. Viburg II did not address the issue now before us. Nor did we suggest in that case that we were overruling or retreating from Medina. The only issue before us in Viburg II was whether double jeopardy and due process precluded a defendant's retrial when his felony DUI conviction was reversed on direct appeal. Viburg II, ¶ 10, 500 P.3d at 1127. To the extent that Viburg II addressed Medina, it merely pointed out why that case was distinguishable. See id. at ¶¶ 25-27, 500 P.3d at 1129-30. And to the extent that we said anything more about Medina, our discussion was dicta. See id.
¶98 For all of these reasons, I would conclude that the trial court committed structural error when it sentenced Crabtree to felony DUI and, thus, that conviction must be vacated.
B. Time to Assess Plain Error
¶99 The majority next declines to follow settled Supreme Court case law providing that appellate courts must assess whether a trial court plainly erred by looking to the law in effect at the time of appeal. In so ruling, the majority asserts that, unlike in the federal courts, applying a time-of-trial rule causes no unfairness to criminal defendants in Colorado. Maj. op. ¶¶ 58-59. The majority further concludes that Henderson v. United States, 568 U.S. 266 (2013), which reiterated the time-of-appeal rule that the Supreme Court had adopted in Johnson, purportedly did so conditioned on the applicability of the fourth prong of the federal plain error standard, which we have not adopted. Maj. op. ¶¶ 6, 63-67. And the majority says that principles of stare decisis preclude us from embracing a time-of-appeal rule. Id. at ¶ 67. I disagree with all three points.
¶100 Regarding the concern for unfairness, I believe that it is manifestly unfair to allow a Colorado defendant who happens to be tried and convicted the week before Linnebur was decided and whose trial involved Linnebur error to face a felony conviction, while an identical defendant who is tried and convicted the week after Linnebur was decided and whose trial involved the same Linnebur error would face only a misdemeanor conviction (because, assuming plain error review applies, only the latter defendant could take advantage of the change in the law).
¶101 Nor do I agree with the majority that Crim. P. 35(c)(1) mitigates any possible unfairness. Maj. op. ¶ 59. Crim. P. 35(c)(1) provides that if a criminal defendant files a timely appeal of a conviction and if the judgment has not yet been affirmed on appeal, then that defendant "may file an application for postconviction review upon the ground that there has been a significant change in the law, applied to the applicant's conviction or sentence, allowing in the interests of justice retroactive application of the changed legal standard." On its face, this provision is narrow. It allows a defendant to file a motion for postconviction relief while an appeal is still pending, but it does not require the postconviction court to grant any relief. Rather, it appears to afford the postconviction court discretion to determine whether the interests of justice support the retroactive application of the changed legal standard, and I am less optimistic than the majority appears to be that this rule will prove to mitigate the unfairness that I believe is likely to result from the rule that the majority applies today.
¶102 Regarding the majority's view that Henderson was somehow conditioned on the applicability of the fourth prong of the federal plain error standard, the Supreme Court's case law demonstrates otherwise. The Court first addressed the question of the time to assess plain error in Johnson, 520 U.S. at 468. In that case, which involved a scenario in which the law had been settled at the time of trial but changed by the time of direct appeal, the Court observed that its plain error precedent had refrained from deciding when an error must be plain to be reviewable. Id. at 464, 468. The Government asserted that for an error to be plain, it had to have been plain both at the time of trial and the time of appeal. Id. at 467. The defendant, in contrast, argued "that such a rule would result in counsel's inevitably making a long and virtually useless laundry list of objections to rulings that were plainly supported by existing precedent." Id. at 468. The Court, with eight Justices joining the opinion, agreed with the defendant and held that "in a case such as this- where the law at the time of trial was settled and clearly contrary to the law at the time of appeal-it is enough that an error be 'plain' at the time of appellate consideration." Id. At no point did the Court condition this ruling on the applicability of the fourth prong of the federal plain error test.
¶103 The issue of the time at which plain error is to be assessed came before the Court again in Henderson, 568 U.S. at 268-69. That case required the Court to decide whether the time-of-appeal plain error standard that the Court had announced in Johnson extended to the scenario in which a legal question was unsettled at the time of trial but settled by the time of appeal. Id. The Court concluded that it did: "In our view, as long as the error was plain as of that later time-the time of appellate review-the error is 'plain' within the meaning of the [Plain Error] Rule." Id. at 269.
¶104 In reaching this conclusion, the Court began by noting that parties may forfeit their rights by failing to assert them on a timely basis. Id. at 271. The Court observed, however, that "[t]he general rule . . . is that an appellate court must apply the law in effect at the time it renders its decision." Id. (alterations in original) (quoting Thorpe v. Hous. Auth. of Durham, 393 U.S. 268, 281 (1969)); accord People v. Owens, 2024 CO 10, ¶ 112, 544 P.3d 1202, 1228; see also People v. Stellabotte, 2018 CO 66, ¶ 3, 421 P.3d 174, 175 (holding that ameliorative, amendatory legislation applies to convictions pending on direct appeal, unless the amendment contains language indicating that it applies only prospectively). The Court thus noted that, as Chief Justice Marshall wrote long ago:
It is in the general true that the province of an appellate court is only to enquire whether a judgment when rendered was erroneous or not. But if subsequent to the judgment and before the decision of the appellate court, a law intervenes and positively changes the rule which governs, the law must be obeyed, or its obligation denied....
In such a case the court must decide according to existing laws, and if it be necessary to set aside a judgment, rightful when rendered, but which cannot be affirmed but in violation of law, the judgment must be set aside.Henderson, 568 U.S. at 271 (alteration in original) (quoting United States v. Schooner Peggy, 5 U.S. (1 Cranch) 103, 110 (1801)).
¶105 The Henderson Court further observed that, given Johnson, a time-of-error interpretation "would prove highly, and unfairly, anomalous." Id. at 273. On this point, the Court opined that everyone would agree that an error would be "plain" as long as the trial court's decision was plainly incorrect at the time that it was made. Id. The Court further pointed out that pursuant to Johnson, an error would be plain even if the trial court was plainly correct at the time it ruled, if a subsequent change in the law rendered the court's ruling incorrect at the time of appeal. Id. at 273-74. The Court then asked, rhetorically:
[I]f the Rule's words "plain error" cover both (1) trial court decisions that were plainly correct at the time when the judge made the decision and (2) trial court decisions that were plainly incorrect at the time
when the judge made the decision, then why should they not also cover (3) cases in the middle-i.e., where the law at the time of the trial judge's decision was neither clearly correct nor incorrect, but unsettled?Id. at 274. In the Court's view, "[t]o hold to the contrary would bring about unjustifiably different treatment of similarly situated individuals." Id.
¶106 The Court also noted that applying the law in effect at the time of trial would require appellate courts to "play a kind of temporal ping-pong." Id. at 275. An appellate court would need to consider the law at the time of appeal to determine whether an error exists, then look at the law at the time of trial to conclude whether the error was plain, and finally weigh the current circumstances to decide whether the error was substantial. Id. The Court disapproved of this approach, reasoning that it would create a more complex and time-consuming appellate process. Id.
¶107 On this point, the Court rejected the Government's contention that a time-of-error rule would properly incentivize counsel to call the matter to the trial court's attention at a time when the judge could take remedial action. Id. at 275-76. The Court observed, "[I]n the present context, any added incentive has little, if any, practical importance." Id. at 276. The Court then reiterated that a time-of-review interpretation furthers the basic principle requiring an appellate court to apply the law in effect at the time the court renders its decision. Id. In contrast, a time-of-error rule "is out of step with [the Court's] precedents, creates unfair and anomalous results, and works practical administrative harm." Id. at 276-77.
¶108 Lastly, the Court noted that Johnson had made clear that "plain-error review is not a grading system for trial judges." Id. at 278. Instead, plain-error review has "broader purposes," including allowing appellate courts to identify those instances in which applying the law that exists at the time of appeal will ensure fairness and judicial integrity. Id.
¶109 In my view, the same principles apply with equal force here. Applying the law at the time of appeal is consistent with centuries of Supreme Court case law, as well as precedent of this court, and I do not see any persuasive reason to deviate from that rule now. Such a rule also promotes fairness, the interests of justice, and judicial integrity, and it avoids "unjustifiably different treatment of similarly situated individuals." Id. at 274. Accordingly, applying the law in effect at the time of appeal would safeguard a defendant's rights and properly prioritize the rule of law over a trial court's misplaced concern that a determination of plain error somehow amounts to a poor evaluation of that court's work.
¶110 Notwithstanding the majority's assertion to the contrary, Maj. op. ¶¶ 6, 63-67, nothing in either Johnson or Henderson conditioned a time-of-appeal rule on the fourth prong of United States v. Olano, 507 U.S. 725, 736 (1993), which we have at least implicitly rejected by continuously applying our three-pronged test for plain error review, see, e.g., Cardman v. People, 2019 CO 73, ¶ 19, 445 P.3d 1071, 1079 (stating that "plain error occurs when there is (1) an error, (2) that is obvious, and (3) that so undermines the fundamental fairness of the trial itself as to cast serious doubt on the reliability of the judgment of conviction"); Miller, 113 P.3d at 750 (same), and which the majority expressly rejects today, Maj. op. ¶¶ 68-71. To the contrary, as noted above, the focus of both Johnson and Henderson was on the age-old principle that appellate courts apply the law in effect at the time of direct review.
¶111 To the extent that Henderson mentioned Olano, it was merely to assuage the prosecution's concern that applying a time-of-appeal rule would open the "plain error" floodgates. Henderson, 568 U.S. at 278-79. In reassuring the prosecution that such a rule would not do so, the Court observed that not every change in a rule of law will render all contrary determinations by trial courts plainly erroneous. Id. at 278. Moreover, in instances in which the law has changed, the Court reasoned that the third and fourth prongs of Olano (comprising the requirement that an error affect the defendant's substantial rights) have served as a backstop against non-meritorious plain error claims. Id. at 278-79.
¶112 This same reasoning applies regarding current Colorado law. It is not all that common for the law to change between the time of trial and the time of appeal, and a change in the law does not mean that all prior contrary determinations are plainly erroneous. Specifically, the third prong of our own plain error test (i.e., our requirement that an error affect the defendant's substantial rights) provides precisely the same backstop as does the federal test: if an error did not affect a defendant's substantial rights, then an appellate court will conclude that the error was not plain.
¶113 Finally, I am unpersuaded by the majority's view that stare decisis precludes us from embracing a time-of-appeal rule. Maj. op. ¶ 67. Under settled stare decisis principles, we may depart from prior precedent when we are convinced that "(1) [our prior] rule was originally erroneous or is no longer sound because of changing conditions and (2) more good than harm will come from departing from precedent." Love v. Klosky, 2018 CO 20, ¶ 15, 413 P.3d 1267, 1270. Here, it is not at all clear to me that the time-of-trial rule is as settled as the majority presumes. As the majority notes, Maj. op. ¶ 52, we have granted certiorari on a number of occasions to decide whether to embrace the time-of-appeal rule, suggesting to me that the issue has been unsettled for some time. Even if the time-of-trial rule were settled, however, for the reasons discussed above, I believe that that rule was originally erroneous because it is inconsistent with centuries of settled law indicating that on direct appeal, an appellate court must apply the law in effect at the time it renders its decision. Moreover, under the principles that we expressed in People v. Novotny, 2014 CO 18, ¶¶ 17-26, 320 P.3d 1194, 1200-03, the Supreme Court's evolving view of the law regarding when to assess plain error has signaled the changing conditions warranting our departure from prior precedent. And because, for the reasons that I have discussed, a time-of-appeal rule promotes fundamental fairness, more good than harm will come from our embracing that rule.
¶114 In light of the foregoing, I would conclude that appellate courts should assess plain error by looking at the law in effect at the time of appeal, rather than the law in effect at the time of trial.
¶115 Accordingly, even if the error that occurred in this case were somehow not structural, I believe that Crabtree would be entitled to the benefit of our ruling in Linnebur, and as the division below concluded, his felony conviction should be vacated.
III. Conclusion
¶116 For these reasons, I would conclude that the trial court structurally erred when it elevated Crabtree's misdemeanor DUI conviction to a felony DUI conviction, despite the jury's not having found the facts of Crabtree's prior convictions beyond a reasonable doubt and the trial court's having made such findings by a preponderance of the evidence. For this reason alone, I would conclude that Crabtree's felony DUI conviction should be vacated.
¶117 Even if the foregoing error were not structural, however, I would conclude that any review for plain error must consider the law in effect at the time of appeal, consistent with the long-settled principle that appellate courts apply the law in effect at the time of direct review. Because Linnebur provided the applicable rule of decision at the time of Crabtree's direct appeal, I would conclude that any review for plain error would likewise require that Crabtree's felony conviction be vacated.
¶118 Accordingly, I would affirm the judgment of the division below, albeit on somewhat different grounds, and I therefore respectfully dissent.