(dissenting). Once again, the majority “swerves and dodges the decisions of the United States Supreme Court” to deny indigent defendants access to justice, this time by refusing to retroactively apply the rule of Halbert v Michigan, 545 US 605; 125 S Ct 2582; 162 L Ed 2d 552 (2005). See People v Bulger, 462 Mich 495, 522; 614 NW2d 103 (2000) (CAVANAGH, J., dissenting). Ironically, the majority now applies the same reasoning that the United States Supreme Court rejected in Halbert to conclude that Halbert should not apply retroactively. I must respectfully dissent.
Before 1994, indigent defendants in Michigan who had pleaded guilty could appeal as of right and were commonly provided with appellate counsel. See, e.g., People v Ginther, 390 Mich 436, 444; 212 NW2d 922 (1973). In 1994, the Michigan Constitution was amended to provide that appeal from a guilty or nolo contendere plea was “by leave of the court.” Const 1963, art 1, § 20. Some trial courts began to interpret this amendment as abolishing the right to counsel for indigent defendants who had pleaded guilty. Halbert, 545 US at 609. This interpretation was codified by the Legislature in MCL 770.3a, which stated that defendants who plead guilty “shall not have appellate counsel appointed for review,” with certain exceptions.
A majority of this Court upheld the constitutionality of MCL 770.3a in People v Harris, 470 Mich 882 (2004) *404(relying on the analysis of Bulger, supra). The majority came to this conclusion by reasoning that first-tier review of plea-based convictions is discretionary, that plea proceedings are simpler than proceedings at trial, and that a defendant who enters a guilty plea accedes to the state’s interest in the finality of criminal proceedings. Bulger, 462 Mich at 508, 516-517; see also Halbert, 545 US at 613-614. I dissented, urging the United States Supreme Court to “correct the constitutional miscarriage committed by the majority” and to “issue the decision that is uniformly directed by its past opinions.” Bulger, 462 Mich at 522-523 (CAVANAGH, J., dissenting).
The United States Supreme Court did indeed correct the error of Bulger and Harris in its 2005 Halbert decision. There the Court held that MCL 770.3a was unconstitutional and restored the constitutional right to the appointment of counsel for first-tier appellate review for indigent defendants in Michigan who had pleaded guilty. Now the question is whether indigent defendants whose plea-based convictions became final between 1994 and 2005 should have the constitutional relief Halbert demands by retroactive application of that decision. The majority arbitrarily cuts off constitutional relief to these indigent defendants, applying the same faulty reasoning it used to deny their constitutional rights in the first place.
The majority concludes that Halbert should not apply retroactively under Teague v Lane, 489 US 288; 109 S Ct 1060; 103 L Ed 2d 334 (1989). I observe first that Teague is inapplicable to this case. The United States Supreme Court has stated that “[a] close reading of the Teague opinion makes clear that the rule it established was tailored to the unique context of federal habeas and therefore had no bearing on whether States could *405provide broader relief in their own postconviction proceedings than required by that opinion.” Danforth v Minnesota, _ US _; 128 S Ct 1029, 1039; 169 L Ed 2d 859 (2008) (emphasis added). The Court went on to say that “[i]t is a matter that States should be free to evaluate, and weigh the importance of, when prisoners held in state custody are seeking a remedy for a violation of federal rights by their lower courts.” Id. at _, 128 S Ct at 1041. Thus, Teague does not restrain this Court from providing a remedy that it previously wrongfully denied.
Nonetheless, I believe that even Teague counsels retroactive application in this case. Teague held that, generally, courts should not retroactively apply rules of criminal procedure that are “new.” The rule of Halbert is not new. First, and most obviously, the rule of Halbert is not new because it reinstated an old rule. See, e.g., Ginther, supra. Halbert merely restores the law that existed in Michigan before 1994. Thus, it is not new.
A rule may be new, under Teague, if “the result was not dictated by precedent existing at the time the defendant’s conviction became final.” Teague, 489 US at 301 (opinion of O’Connor, J.). As I stated in Bulger, I believe that the result of Halbert was uniformly directed by the past decisions of United States Supreme Court. Bulger, 462 Mich at 522-523 (CAVANAGH, J., dissenting). A rule that is uniformly directed is not new.
The majority opinion concludes that the rule of Halbert is not compelled, and thus new, because appeal from a guilty plea in Michigan is by leave and discretionary. It reasons that the application of Douglas v California, 372 US 353; 83 S Ct 814; 9 L Ed 2d 811 (1963), was not dictated because Douglas involved a first appeal as of right, while the first appeal of plea-based convictions in Michigan is discretionary. The *406majority opinion further reasons that Ross v Moffitt, 417 US 600; 94 S Ct 2437; 41 L Ed 2d 341 (1974), appears to apply because it held that appointment of counsel was not required when appeal is made to a court that has discretion “to choose whether to reach the merits of a defendant’s appeal.” Ante at 389. First, I observe that this very reasoning was rejected by the United States Supreme Court in Halbert. The Court stated that “Halbert’s case is properly ranked with Douglas rather than Ross” and, thus, held “that the Due Process and Equal Protection Clauses require the appointment of counsel for defendants, convicted on their pleas, who seek access to first-tier review in the Michigan Court of Appeals.” Halbert, 545 US at 610 (emphasis added).
Second, the fact that the defendant’s appeal in Douglas was as of right was irrelevant to the outcome of that case. Rather, the critical issue was the fact that, “[wjhether formally categorized as the decision of an appeal or the disposal of a leave application, the Court of Appeals’ ruling on a plea-convicted defendant’s claims provides the first, and likely the only, direct review the defendant’s conviction and sentence will receive.” Halbert, 545 US at 619. In Ross, the discretion involved was irrespective of the merits. As the Ross Court stated, its ruling applied to appellate courts that may deny leave even when they conclude that the decision on the merits in the court below was incorrect. Ross, 417 US at 615. That is not the case when a defendant seeks first-tier review in the Michigan Court of Appeals. See Bulger, 462 Mich at 541-542 (CAVANAGH, J., dissenting); Halbert, 545 US at 617 (“Michigan’s intermediate appellate court looks to the merits of the claims made in the application”). Further, Ross made clear that its decision applied when a defendant had already “received the benefit of counsel in examining *407the record of his trial and in preparing an appellate brief on his behalf for the state Court of Appeals” and when “a defendant’s claims of error are organized and presented in a lawyerlike fashion to the Court of Appeals .. . Ross, 417 US at 614-615. That is not the case when a defendant seeks first-tier review in the Michigan Court of Appeals.
Thus, this precedent does not support a claim that a reasonable jurist could conclude that the rule of Halbert was not compelled. To the contrary, “[t]he Michigan Supreme Court’s reading [of] Ross to permit the denial of counsel to an indigent defendant on appeal solely because the appeal is discretionary [is] not a reasonable application of Supreme Court precedent.” Bulger v Curtis, 328 F Supp 2d 692, 702 (ED Mich, 2004). Because I believe that the rule of Halbert was, in fact, compelled by precedent, I believe that the rule is not new. Thus, Halbert should apply retroactively. Even if the rule of Halbert were new, it would represent a “watershed” decision, which requires retroactive application under Teague. Teague states that “a new rule should be applied retroactively if it requires the observance of those procedures that. .. are implicit in the concept of ordered liberty.” Teague, 489 US at 307, 311 (opinion of O’Connor, J.) (citations omitted). At issue here is meaningful access to the courts, Ross, 417 US at 615, and the essential fairness of state-ordered proceedings, Halbert, 545 US at 611. I believe that these are matters that are “implicit in the concept of ordered liberty.”
The majority supports its finding to the contrary with the assertion that a state is not required to provide “any appellate proceedings at all for defendants who plead guilty.” Again, the United States Supreme Court rejected this reasoning in Halbert. The Court reminded *408that while “[t]he Federal Constitution imposes on the States no obligation to provide appellate review of criminal convictions,” nonetheless, once provided, “a State may not bolt the door to equal justice to indigent defendants.” Halbert, 545 US at 610 (internal citations and quotation marks omitted). I conclude that the rule of Halbert fits this exception to Teague. It should apply retroactively.
This conclusion is supported by the fact that Douglas, the case on which the United States Supreme Court based its Halbert decision, was applied retroactively. McConnell v Rhay, 393 US 2, 3; 89 S Ct 32; 21 L Ed 2d 2 (1968). Observing this application, the Supreme Court grouped Douglas with Gideon v Wainwright, 372 US 335; 83 S Ct 792; 9 L Ed 2d 799 (1963), as cases implicating the right to counsel and addressing a right that “relates to the very integrity of the fact-finding process.” McConnell, 393 US at 3 (internal citation and quotation marks omitted). Thus, the majority opinion’s attempt to distinguish this case from the sort that announces a “watershed” rule is incorrect. Ante at 391-392. Douglas was decided on equal protection and due process grounds, just like Halbert. Yet the Supreme Court identified Douglas as implicating the same right as Gideon under a different constitutional provision. Thus, Douglas would suggest that, under Teague, Halbert is a “watershed” rule requiring retroactive application irrespective of the specific constitutional ground on which it was decided.
But, as noted, Teague does not control the measure of retroactivity applied by a state court. Rather, Michigan jurisprudence provides the tools for assessment in this case.1 The majority applies the factors stated in People *409v Sexton, 458 Mich 43, 580 NW2d 404 (1998), to conclude that retroactivity is not required in this case. I believe that the Sexton factors direct the opposite result.
The first Sexton factor, the purpose factor, states that a law may be applied retroactively when it “concerns the ascertainment of guilt or innocence”; however, “a new rule of procedure ... which does not affect the integrity of the fact-finding process should be given prospective effect.” Id. at 63 (citations and quotation marks omitted). The majority concludes that this factor is inapplicable because “the appointment of counsel on appeal does not concern the ascertainment of guilt or innocence.” Ante at 393-394.1 strongly disagree. I believe the ascertainment of guilt or innocence is at stake here because “ 'a correct adjudication of guilt’ involves more than just an admission of guilt.” Bulger, 462 Mich at 560 (CAVANAGH, J., dissenting); see also MCR 6.302. “Appeals after guilty pleas, too, directly implicate a procedure without which the accuracy of a conviction cannot be assured.” Simmons v Kapture, 516 F3d 450, 457 (CA 6, 2008) (Martin, J., dissenting), citing Halbert, 545 US at 617. Halbert rests precisely on the fact that a defendant’s first-tier appeal from a plea-based conviction involves error-correction. In other words, a defendant’s guilt or innocence is at stake. Thus, the purpose prong directs retroactive application of Halbert.
I observe further that the majority opinion relies on the presumption that all defendants who plead guilty are indeed guilty. Ante at 394. As I pointed out in Bulger, this misses the entire purpose of a first-tier appeal from a guilty plea, where factors relevant to *410guilt — such as coercion, ineffective assistance of counsel, and mental capacity — are meant to be adjudicated.2 Correcting these errors is relevant precisely to the question of guilt or innocence. In McConnell, supra, the United States Supreme Court stated the significance of the issue at stake:
This Court’s decisions on a criminal defendant’s right to counsel at trial, Gideon v. Wainwright, 372 US 335 (1963); at certain arraignments, Hamilton v. Alabama, 368 US 52 [82 S Ct 157; 7 L Ed 2d 114] (1961); and on appeal, Douglas v. California, 372 US 353 (1963), have been applied retroactively. The right to counsel at sentencing is no different. As in these other cases, the right being asserted relates to the very integrity of the fact-finding process. [McConnell, 393 US at 3 (citations and quotation marks omitted).]
The Supreme Court in Halbert observed that the error-correction function of a first-tier review in the Michigan Court of Appeals was crucial to its conclusions in that case. Halbert, 545 US at 617. Where the very integrity of the fact-finding process is at stake, retroactive application is directed. Further, the majority’s position here is another way of stating that a defendant who pleads guilty accedes to the state’s interest in finality, a proposition the United States Supreme Court rejected in Halbert. Id. at 623-624.
*411The second Sexton factor is also implicated. This factor addresses the “general reliance on the old rule.” Sexton, 458 Mich at 60. Addressing this factor, the Sexton Court stated that “[j]udicial decisions are generally given complete retroactive effect unless the decisions are unexpected or indefensible.” Id. at 63-64, citing People v Doyle, 451 Mich 93, 104; 545 NW2d 627 (1996). As exhaustively demonstrated in Halbert, this rule is neither “unexpected” nor “indefensible.”
In this case, injustice will result if Halbert is not applied retroactively. The majority’s decision in Bulger left indigent defendants who pleaded guilty with a “meaningless ritual in our Court of Appeals.” Bulger, 462 Mich at 581 (CAVANAGH, J., dissenting). Failure to apply Halbert retroactively means that for a “small group of people arbitrarily caught between Michigan’s own protections [before 1994] and the protection offered by Halbert, the ‘meaningless ritual’ of indigent appeals continues to be a harsh and unjust reality ....” Simmons, 516 F3d at 458 (Martin, J., dissenting). As the United States Supreme Court stated in Douglas, the case on which it relied for the rule of Halbert, “When society acts to deprive one of its members of his life, liberty or property, it takes its most awesome steps.” Douglas, 372 US at 358 n 2 (citation and quotation marks omitted). Unless Halbert is applied retroactively, defendants whose plea-based convictions became final during the arbitrary period between 1994 and 2005 will be penalized by the general reliance on an unconstitutional ruling of this Court. The second factor of Sexton directs retroactive application of Halbert.
Finally, the effect on the administration of justice, the third Sexton factor, requires retroactive application. The very system of justice administered by this Court rests on the fair application of fundamental rights, such *412as the right to counsel on first-tier appellate review. As the Supreme Court observed in Douglas:
“No general respect for, nor adherence to, the law as a whole can well be expected without judicial recognition of the paramount need for prompt, eminently fair and sober criminal law procedures. The methods we employ in the enforcement of our criminal law have aptly been called the measures by which the quality of our civilization may be judged.” [Douglas, 372 US at 358 n 2, quoting Coppedge v United States, 369 US 438, 449; 82 S Ct 917; 8 L Ed 2d 21 (1962).]
The majority concludes that the effect on the administration of justice counsels against retroactivity because “[t]he state’s strong interest in finality of the criminal justice process would be undermined.” Ante at 397. In Halbert, the United States Supreme Court found this argument insufficient to support the denial of appellate counsel to defendants who had pleaded guilty. Halbert, 545 US at 623. Further, the majority’s contention that retroactive application oí Halbert would “inundate the appellate process with new appeals” is speculative. Ante at 398. As the majority observes, only a small percentage of defendants who pleaded guilty availed themselves of appointed counsel to seek an appeal before such appointment was precluded by the unconstitutional rule of MCL 770.3a. Ante at 396-397. Also, appointed counsel would be prohibited from asserting frivolous claims for appeal by Rule 3.1 of the Michigan Rules of Professional Conduct. And the Court of Appeals retains discretion on whether to grant leave to hear appeals from guilty pleas.3 Const 1963, art 1, § 20. Finally, as the majority also observes, a defendant *413must consider the risk that an adverse ruling on appeal may result in a more severe penalty. Ante at 395; MCR 6.312. Together, these factors suggest that only a small number of cases with genuine and substantial issues for appeal will receive full consideration by our state appellate courts. But even if the number of appeals would be great, as the majority speculates, defendants validly asserting claims of substantial error should be heard in our appellate courts.
It strikes me as an ironic twist to apply the invalid reasoning that the majority originally used in attempting to justify denying these defendants their constitutional right to now deny them review retroactively. I see no reason to deny constitutional rights to defendants on the arbitrary basis that their convictions became final between 1994 and 2005. On the contrary, I believe that failure to apply the rule of Halbert retroactively is unreasonable and constitutionally unconscionable. I must respectfully dissent.
KELLY, J., concurred with CAVANAGH, J.I am aware of the decision in Simmons v Kapture, 516 F3d 450 (CA 6, 2008). I am also aware that the petition for writ of certiorari was denied *409in Houlihan v Michigan, _ US _; 129 S Ct 254; 172 L Ed 2d 191 (2008). As noted earlier, these decisions do not prohibit or affect the application of state law to this case.
In greater detail, I stated that appeal from a plea-based conviction may involve
constitutional defects that are irrelevant to [a defendant’s] factual guilt, double jeopardy claims requiring no further factual record, jurisdictional defects, challenges to the sufficiency of the evidence at the preliminary examination, preserved entrapment claims, mental competency claims, factual basis claims, claims that the state had no right to proceed in the first place, including claims that a defendant was charged under an inapplicable statute, and claims of ineffective assistance of counsel. [Bulger, 462 Mich at 561 (Cavanagh, J., dissenting) (citations omitted).]
Of course, as an error-correcting court, the Court of Appeals must conduct some analysis of each application for leave to appeal before exercising its discretion on whether to hear each case. This analysis is evidently less burdensome than actually hearing an appeal. Additionally, *413the Court would have the advantage of reviewing arguments that have been researched and briefed by counsel.