dissenting:
The majority has ruled that a lawyer fails to render the effective assistance of counsel guaranteed by the United States Constitution when he does not pursue before a state’s highest court a state law claim rejected on the merits by a state’s intermediate appellate court. Because I am unable to agree with that unprecedented ruling, I respectfully dissent.
Angel Claudio has twice received less than top-flight legal representation in the course of proceedings that led to his state court conviction for second-degree murder. First, his state court trial counsel failed to advise him fully before he confessed to the crime. That failure was arguably a violation of both his Sixth Amendment right to the effective assistance of counsel and his broader state law right protected by the New York constitution. The state trial court suppressed his confession on Sixth Amendment grounds, but the Appellate Division, on an interlocutory appeal by the prosecution, properly reversed the suppression order, since the confession had been made before the Sixth Amendment right to counsel had attached. People v. Claudio, 85 A.D.2d 245, 447 N.Y.S.2d 972 (2d Dep’t 1982).
Though Claudio’s state court appellate counsel had defended the suppression order only on Sixth Amendment grounds, the Appellate Division, on its own motion, elected to consider whether the confession had been obtained in violation of state law. The Appellate Division ruled that Claudio had a state law right to counsel when he confessed, id. at 256, 447 N.Y.S.2d at 979, but that this right assured him only the presence of counsel, rather than performance of counsel above some minimum level of professional competence, id. at 257-61, 447 N.Y.S.2d at 980-82. In the Appellate Division’s view, the quality of counsel’s representation, as a matter of state law, must meet standards of adequacy when rendered in court where judicial supervision is available, but standards of adequacy cannot be enforced with respect to legal services rendered in investigative proceedings where judicial supervision is less readily available. Id.
At that point, the second lawyering inadequacy occurred. Challenging the Appellate Division’s reversal of the suppression order in an appeal to the New York Court of Appeals, Claudio’s state court appellate counsel argued only that trial counsel’s performance had violated Claudio’s federal rights under the Sixth Amendment and failed to include in his appeal the claim that the confession violated Claudio’s state law right to effective assistance of counsel. Though the state law claim had been adjudicated by the Appellate Division to be without merit, a majority of the panel hearing this habeas corpus appeal now concludes that the state law claim had sufficient likelihood of success that Claudio’s state court appellate counsel denied him a federally protected right to the effective assistance of counsel by failing to argue the state law claim to the Court of Appeals.
This ruling unduly extends habeas corpus jurisdiction and the Due Process Clause. To reach such a ruling, the majority takes several steps that need to be separately identified.
*8071. “Critical stage. ” The majority holds that the appeal to the New York Court of Appeals from the Appellate Division was a “critical stage,” see United States v. Wade, 388 U.S. 218, 227, 87 S.Ct. 1926, 1932, 18 L.Ed.2d 1149 (1967), at which Claudio was entitled under the Due Process Clause to the effective assistance of counsel. I have no doubt that this appeal was an important stage, the outcome of which could be critical for Claudio. But whether it was a “critical stage” for Sixth Amendment purposes is uncertain.
The Supreme Court has ruled that whatever right to counsel exists at the post-conviction appellate stage of a state criminal case is secured directly by the Due Process Clause of the Fourteenth Amendment, rather than through incorporation of the Sixth Amendment, which protects the “trial-level right to counsel,” Evitts v. Lucey, 469 U.S. 387, 392, 105 S.Ct. 830, 834, 83 L.Ed.2d 821 (1985); see Ross v. Moffitt, 417 U.S. 600, 608-11, 94 S.Ct. 2437, 2442-44, 41 L.Ed.2d 341 (1974); Miller v. Keeney, 882 F.2d 1428, 1431 n. 4 (9th Cir.1989).1 The Court has also ruled that there is no Due Process right to counsel in discretionary post-conviction appeals. See Coleman v. Thompson, — U.S. —, —, 111 S.Ct. 2546, 2568, 115 L.Ed.2d 640 (1991); Wainwright v. Torna, 455 U.S. 586, 587-88, 102 S.Ct. 1300, 1301-02, 71 L.Ed.2d 475 (1982). Claudio’s appeal to the Court of Appeals was discretionary, but it was a pre-conviction appeal. The majority relies on Ross v. Moffitt, 417 U.S. at 610-11, 94 S.Ct. at 2443-44, where the Court noted the significant difference between having counsel prior to conviction “as a shield” and having counsel after conviction “as a sword.” Id. I am not certain whether the Supreme Court would afford Sixth Amendment protection to a discretionary pre-trial appeal to challenge the lawfulness of a confession when such protection would not be available directly under the Due Process Clause to challenge the same confession on a discretionary post-trial appeal. But I will assume, for the argument, that the Sixth Amendment attached to the pre-trial discretionary appeal and consider the majority’s subsequent steps.2
2. Obligation to raise state law issues. The majority next rules that a lawyer’s failure to pursue on a state court appeal a state law claim can constitute a denial of federally guaranteed effective assistance of counsel. That is a significant ruling, not yet made by the Supreme Court,3 nor by any prior decision of this Court. It is arguable that the federal guarantee of effective assistance of counsel, whether protected by the Sixth Amendment through the Due Process Clause or directly by that Clause, *808assures state court defendants only effective assistance of counsel in enforcing the protections of federal law, leaving entirely to the states all questions as to whether counsel’s performance was adequate to assure vindication of state law protections.
However, the few cases that have considered the issue, either explicitly or implicitly, have all ruled, or at least assumed, that counsel may be found ineffective for purposes of the federal Constitution by failing to pursue a state law claim in state court. Counsel was held ineffective for omitting a state law issue in Nero v. Blackburn, 597 F.2d 991, 993-94 (5th Cir.1979); see also Laffosse v. Walters, 585 F.Supp. 1209, 1212-13 (S.D.N.Y.1984) (semble) (counsel ineffective for filing Anders brief where non-frivolous claims, including state law claims, available), and in other cases, the adequacy of counsel has been assessed with reference to state law issues, though counsel was not deemed deficient. See Miller v. Keeney, 882 F.2d 1428, 1435 (9th Cir.1989); Parton v. Wyrick, 704 F.2d 415, 416-17 (8th Cir.1983).
Though the matter is not free from doubt, I agree with the majority that the federally guaranteed right to the effective assistance of counsel includes some obligation of counsel to assert claims and defenses grounded in state as well as federal law.
3. Standard for assessing competence of appellate counsel. The majority next considers the standard to be applied in determining whether counsel’s failure to present a particular state law claim on appeal violates the federally protected right to effective assistance of counsel. The majority rules that the standard for assessing the competency of counsel in the trial context, see Strickland v. Washington, 466 U.S. 668, 104 S.Ct. 2052, 80 L.Ed.2d 674 (1984), applies equally to the appellate context.4 Strickland announced a two-prong test for assessing ineffective assistance of counsel claims in the pre-trial and trial context. The first prong concerns counsel’s performance. The claimant must show that counsel’s performance fell below “an objective standard of reasonableness,” Strickland, 466 U.S. at 688, 104 S.Ct. at 2064, which means “simply reasonableness under prevailing professional norms.” Id. The second prong concerns prejudice. The claimant must show a “reasonable probability” that reasonably competent assistance would have affected the outcome. Id. at 694, 104 S.Ct. at 2068.
I agree that the first prong of Strickland applies fully to the appellate context, but applying the second prong is unnecessary in the appellate context and risks both unfairness to habeas petitioners and excessive involvement of federal courts in state law issues. Though we have previously stated that both prongs of Strickland apply in the appellate context, see Abdurrahman v. Henderson, 897 F.2d 71, 74 (2d Cir.1990), that decision involved no application of the first prong of Strickland — the competency of appellate counsel — since the panel assumed, for purposes of the habeas appeal, that counsel’s performance had been constitutionally deficient. Id. at 74. The objections to applying Strickland’s second prong to the appellate context do not become apparent until a case like the pending one arises where the constitutional adequacy of appellate counsel needs to be assessed.
Strickland’s second prong is unnecessary in the appellate context. The “reasonable probability of affecting the outcome” standard made very good sense in the context of the Strickland case, a challenge to *809counsel’s performance at the investigation and trial stages of a criminal case. Since a trial cannot be reconvened and resumed to afford the defendant an opportunity to overcome counsel’s misstep and secure a verdict unaffected by counsel’s incompetence, the habeas court must make some reasoned assessment of the probability that competent lawyering would have affected the outcome.5 However, in the context of deficient lawyering on appeal, there is no reason to speculate about the probability of a different outcome: the state appellate court remains fully available to adjudicate the claim that appellate counsel failed to present.
Thus, in the appellate context, there is no reason to apply the second prong of Strickland. If the performance of appellate counsel in omitting a state law issue was outside the range of reasonable professional competence, the habeas court can afford the State the opportunity to resume the appeal by granting a conditional writ that orders a new trial unless the state appellate court entertains and, in the absence of a procedural bar, adjudicates the previously omitted claim. Indeed, that is the remedy adopted by the Court in this case.6
It is arguable that some vestige of the second prong of Strickland should remain for the situation where appellate counsel was incompetent for failing to raise what would have been a winning claim but the claim, if sustained, would not have resulted in a reversal. But there is little point in making a separate “reasonable probability of affecting the outcome” analysis, since appellate counsel could not possibly be deemed to have performed outside the range of reasonable professional compe*810tence by omitting an issue that would have been considered harmless error.
Applying the prejudice prong of Strickland to the appellate context is not only unnecessary, but it also risks altering the first prong by importing into the assessment of competency the “reasonable probability” language of the prejudice inquiry. Thus, in the pending case, the majority starts with the second prong and analyzes the issue omitted by the state court appellate counsel to see if it had a “reasonable probability” of success. Since Strickland used “reasonable probability” to measure the likelihood that a trial outcome would have been different (in order to assess prejudice from lawyering already determined to be deficient), it is ill-advised to press this same standard into service for the different task of determining whether counsel’s performance was deficient.
In my view, the proper analysis for assessing a claim of ineffective appellate counsel is to determine initially whether the failure to pursue the omitted issue is outside the range of reasonable professional competence and, if so, whether there is at least a substantial argument that the alleged error would not be deemed harmless. Of course, in making a determination whether appellate counsel’s omission of an issue was outside the range of reasonable professional competence, the habeas court will have to make some assessment of the merits of the omitted issue. Even if the issue can fairly be considered non-frivolous, its omission is not necessarily professional incompetence, in view of counsel’s broad discretion to select from among available claims those thought more likely to be successful. See Jones v. Barnes, 463 U.S. 745, 751-53, 103 S.Ct. 3308, 3312-14, 77 L.Ed.2d 987 (1983).
By invoking in the appellate context the “reasonable probability of affecting the outcome” standard from the trial context, the majority creates the risk, no doubt inadvertently, that some defendants will be denied an opportunity to have presented to state courts some state law issues that reasonably competent counsel would have asserted and that might, after issuance of a conditional writ, prove successful in state court, but that might not be thought by a federal habeas court to have a “reasonable probability of affecting the outcome.” In other words, appellate counsel might, in some instances, be found to have performed below the level of reasonable competence, even though the habeas court is not persuaded that the-omitted issue is a likely winner. In a case like the pending one, where the omitted claim is based on state law, the test should be simply whether the claim had sufficient merit, in view of the other available issues, that reasonably competent counsel would pursue it; the habeas court’s prediction that the state law claim would probably be meritorious in state court is wholly unnecessary and, indeed, inappropriate for a federal court.
4. Omission of state law issue rejected by state intermediate court. That brings me to the majority’s fourth and most questionable step, the ruling that Claudio’s state court appellate counsel was constitutionally deficient in this case. The majority rules that appellate counsel’s failure to carry to the Court of Appeals the state law claim rejected by the Appellate Division was constitutionally deficient. Whatever standard is used to assess the competency of state court appellate counsel, I cannot agree that counsel was constitutionally deficient in omitting that issue from his second-level appeal.
The majority undertakes a detailed analysis of state law and finds a “reasonable probability” that Claudio’s state law claim would have succeeded in the Court of Appeals. In my view, that analysis extends the constitutional right to counsel beyond the authority of a federal habeas court in a case such as this where an intermediate state appellate court has ruled that the state law claim is without merit.
The Appellate Division faced what it regarded as an issue of first impression under New York law. See People v. Claudio, 85 A.D.2d at 257, 447 N.Y.S.2d at 980. Recognizing that under New York law the right to counsel may attach at an earlier time than under federal law and that the state right to counsel had attached at the *811time of Claudio’s confession, the Appellate Division had to decide whether the state law right to counsel includes not only the presence of counsel but also the competent assistance of counsel prior to the initiation of formal judicial proceedings. The Appellate Division ruled that state law does not provide such protection. It reasoned that the right to the competent assistance of counsel could be enforced by courts where counsel acted in the course of judicial proceedings, but that state law imposed no obligation on police or prosecutors to monitor the quality of legal assistance rendered by defense counsel.7 Id. at 257-61, 447 N.Y.S.2d at 980-82.
I do not know whether the Appellate Division’s ruling is a correct interpretation of New York law. A subsequent decision of the New York Court of Appeals, People v. Beam, 57 N.Y.2d 241, 455 N.Y.S.2d 575, 441 N.E.2d 1093 (1982), decided seven months later, lends arguable support to the Appellate Division’s ruling, but is arguably distinguishable, as the majority in this case contends, at 804. But however New York law will ultimately be explicated on Claudio’s state law claim,8 the undeniable fact is that an intermediate appellate court has ruled that the claim lacks merit. That ruling is not merely a precedent from some other litigation of arguable relevance to Claudio’s case, to which a federal court would owe considerable deference if the correctness of the ruling were directly in issue. See West v. American Telephone and Telegraph Co., 311 U.S. 223, 237-38, 61 S.Ct. 179, 183-84, 85 L.Ed. 139 (1940) (decision of state intermediate appellate court on state law issue “not to be disregarded by a federal court unless it is convinced by other persuasive data that the highest court of the state would decide otherwise”). The ruling was rendered in Claudio’s case.9 In my view, an appellate lawyer can never be found to have denied his client the effective assistance of counsel guaranteed by the federal Constitution for electing to omit on appeal to a state’s highest court a state law claim that has been rejected on the merits by a state’s intermediate appellate court, at least in the absence of some contradictory ruling from the state’s highest court prior to the time the second-stage appeal is filed.
In asserting that virtually per se position, I must reckon with the argument that since appellate counsel can be found to be constitutionally ineffective for omitting on a first-stage appeal a state law issue rejected on the merits by a trial court, the same test of competence should apply when counsel omits from a second-stage appeal an issue rejected on the merits by an intermediate appellate court. The superficial *812symmetry of this argument ignores the nature of the appellate function. Appellate courts sit to examine rulings of trial courts and correct what they perceive to be errors. Trial courts, of course, are equally bound to apply the law, but they function in a different environment. Their trial responsibilities oblige them to rule rapidly. Appellate courts enjoy the luxury of time for more reflective decisionmaking. No doubt some trial court rulings represent the result of prolonged and painstaking examination and some appellate court rulings are hasty and ill-considered. But the institutional differences between the levels of court persuade me that in giving content to the federal guarantee of effective assistance of appellate counsel, a federal court may properly consider counsel’s failure to appeal a trial court’s rejection of a state law claim but may not consider the failure to pursue a second-stage appeal of a state law claim rejected on its merits by an intermediate appellate court.
A proper regard for federalism interests reenforces my conclusion. It is already a significant broadening of the Due Process Clause, with a corresponding expansion of federal habeas jurisdiction, to include state law issues in determining whether counsel’s choice of claims and defenses constituted effective assistance. That step is a reasonable accommodation between the Due Process value of assuring defendants effective participation in the adversary process and the federalism interests in limiting federal court intrusion into matters of state criminal law. But once an intermediate state appellate court has rejected a state law claim on the merits, the role of a federal habeas court would be unduly broadened if it undertook to determine whether counsel was incompetent in not challenging that decision before the state’s highest court. Federal habeas corpus courts have no authority to correct what they perceive to be state court errors of state law. They should be equally barred from undertaking to predict the likelihood that a state’s intermediate court will be reversed by a state’s highest court on an issue of state law, yet that is precisely the assessment the majority makes in this case in determining that Claudio’s counsel was incompetent in not pursuing his state law claim despite the rejection by the Appellate Division.
5. The remedy. Having determined that Claudio’s state court appellate counsel was constitutionally deficient for not pursuing in the Court of Appeals the state law ineffective assistance of trial counsel claim, the majority directs the issuance of a writ of habeas corpus, unless the State affords Claudio an opportunity to “present” his state law claim to the New York Court of Appeals. Normally, where a state prisoner persuades a habeas court that he has been denied a federally protected right, he is entitled to a writ conditioned upon a new trial. In this case, though I find no basis to grant habeas relief, I agree with the majority that if appellate counsel’s performance was constitutionally deficient, the proper remedy is to condition the writ only on the State’s obligation to afford Claudio an opportunity to return to the New York Court of Appeals. See Evitts v. Lucey, 469 U.S. at 390, 105 S.Ct. at 832 (writ conditioned on either retrial or reinstatement of appeal); Laffosse v. Walters, 585 F.Supp. at 1214 (writ conditioned on granting of new appeal). That return, however, is only for the opportunity to “present” his state law claim, as the majority is careful to state. The reason for that limited opportunity arises from the unusual procedural circumstances of Claudio’s case.
Initially, the prosecution took an interlocutory appeal to the Appellate Division from the trial court's ruling suppressing the confession on federal constitutional grounds. When that ruling was reversed, Claudio’s counsel appealed the reversal to the New York Court of Appeals. That Court ruled that Claudio’s state law challenge to his confession would not be considered because appellate counsel had not asserted the state law claim in either the trial court, the Appellate Division, or the Court of Appeals. People v. Claudio, 59 N.Y.2d at 560 n. 1, 466 N.Y.S.2d at 272 n. 1, 453 N.E.2d at 501 n. 1. On the pending appeal, the State argues that the Court of Appeals was procedurally barred from considering the state *813law claim because, wholly apart from appellate counsel’s omission of the issue, an appellate court, under state law, can consider on an interlocutory appeal only the grounds on which a trial court’s ruling rested. See People v. Karp, 76 N.Y.2d 1006, 1009, 565 N.Y.S.2d 751, 751, 566 N.E.2d 1156, 1156 (1990); People v. Cade, 74 N.Y.2d 410, 418, 548 N.Y.S.2d 137, 141, 547 N.E.2d 339, 343 (1989). The majority contends that this procedural bar is inapplicable because the Appellate Division decided the merits of the state law claim and the State briefed the issue in the Court of Appeals.
Whether or not the state law claim was proeedurally barred on the interlocutory appeal is a matter of state law. However that procedural issue would have been resolved had appellate counsel pursued the state law claim in the Court of Appeals on the pre-trial appeal, the outcome will not necessarily be the same on the post-trial appeal now afforded by virtue of the conditional writ. See People v. Goodfriend, 64 N.Y.2d 695, 697, 485 N.Y.S.2d 519, 521, 474 N.E.2d 1187, 1188 (1984). Because we cannot be certain whether the Court of Appeals can now reach the merits of Claudio’s state law claim, I agree that, if, contrary to my view, appellate counsel was constitutionally deficient in not arguing the state law claim to the Court of Appeals, the proper remedy is to afford Claudio only the opportunity to “present” the claim, leaving it to the Court of Appeals to decide whether it has authority to adjudicate that claim.
To the extent indicated in the foregoing, I respectfully dissent.
. The right appears to have originally been grounded on the Equal Protection Clause. See Douglas v. California, 372 U.S. 353, 83 S.Ct. 814, 9 L.Ed.2d 811 (1963).
. In considering whether the appeal to the New York Court of Appeals was a "critical stage” for purposes of a constitutional right to counsel, a third variable is arguably pertinent, in addition to whether the appeal is discretionary or of right, and whether it precedes or follows conviction. The third variable is whether the appeal is forced upon the defendant by the State or initiated by the defendant. The argument for considering the appeal to be a critical stage is stronger where the appeal is initiated by the State. In this case, the assessment of this third factor is complicated by the fact that the State initiated the appeal to the Appellate Division, but the defendant thereafter initiated the further appeal to the Court of Appeals.
. Language in Smith v. Murray, 477 U.S. 527, 106 S.Ct. 2661, 91 L.Ed.2d 434 (1986), gives superficial support to the view that reasonable competence includes consideration of state law issues. In determining whether state appellate counsel’s omission of a claim on direct appeal to a state supreme court was so incompetent as to constitute cause that would excuse procedural default, the Court quoted counsel’s subsequent explanation that " ‘Virginia case law would [not] support our position at that particular time,’ ” id. at 531, 106 S.Ct. at 2664, and concluded that the decision not to pursue the claim was not incompetent "[v]iewed in light of Virginia law at the time [counsel] submitted his opening brief to the Supreme Court of Virginia,” id. at 536, 106 S.Ct. at 2667. However, the claim omitted was a federal claim — that admission of psychiatric testimony violated the defendant’s self-incrimination privilege protected by the Fifth and Fourteenth Amendments. Id. at 531, 106 S.Ct. at 2664. Though the matter is in some doubt, the Court appears to have concluded that counsel reasonably omitted a federal claim because prevailing state court decisions had previously rejected it. See id. at 540, 106 S.Ct. at 2669 (Stevens, J., dissenting).
. The Supreme Court has not yet decided this issue. In Evitts the Court ruled only that if the right to counsel applied to an appellate proceeding, the right included the right to effective assistance of counsel, but the Court left open the issue of whether the standards for assessing the competency of counsel at the appellate stage were the same as at the trial stage. Since the State did not challenge the finding of appellate counsel’s incompetency, the Court said that it "need not decide the content of appropriate standards for judging claims of ineffective assistance of appellate counsel. Cf. Strickland v. Washington, 466 U.S. 668, [104 S.Ct. 2052, 80 L.Ed.2d 674] (1984); United States v. Cronic, 466 U.S. 648 [104 S.Ct. 2039, 80 L.Ed.2d 657] (1984).” Evitts, 469 U.S. at 392, 105 S.Ct. at 833. Arguably, the "cf.” cite to Strickland implies that the appellate standard might be different.
. On the facts of Strickland, a death penalty case, it would have been possible to remedy the allegedly deficient performance of trial counsel without making a "reasonable probability of affecting the outcome" prediction since the deficiency occurred at sentencing rather than during the presentation of evidence before the jury. Thus, had counsel’s performance been found deficient, it would have been possible to issue a conditional writ, conditioned upon the State’s affording new counsel an opportunity to take whatever investigatory steps competent counsel should have taken and then to argue to the trial court judge that the death penalty should not be imposed. But the Supreme Court ruled, even in the context of deficient lawyering occurring at sentencing, that the habeas court is to make a prediction as to whether there is a reasonable probability that a conscientious sentencer would have imposed a different sentence. Strickland, 466 U.S. at 695, 104 S.Ct. at 2068. The Court even imposed this prediction task on a habeas court in situations where a state appellate court performs the role of sentencer by independently reweighing the evidence pertinent to the penalty. Id.
Presumably, the Court was concerned that returning the matter for reconsideration by the same sentencer, after proper presentation by counsel, would not provide an adequate substitute for the decisionmaking that would have occurred if counsel had properly presented defendant’s sentencing arguments at the initial sentencing. Yet, if the habeas court finds a reasonable probability that proper lawyering would probably have affected the sentencing, it is unlikely that the ultimate remedy would be anything more than resentencing. Thus, in Strickland, had counsel been considered deficient and had prejudice been found, the only consequence of the "probability” analysis, compared to omitting such analysis and returning the matter directly to the original sentencer, would have been resentencing before a different judge.
In any event, Strickland’s use of the "probability of affecting the outcome” test, even for lawyering deficiencies occurring at sentencing, does not require a habeas court to make a similar prediction of likely success when the alleged deficiency is the failure to present a legal claim to a state appellate court. Unlike sentencing, which involves an aggregate weighing of a variety of factors, consideration of the omitted issue by the appellate court can readily be performed as effectively upon presentation by new counsel on a new or reinstated appeal as it could have been by the original counsel. Since the Supreme Court, subsequent to Strickland, without pausing to make a "probability of affecting the outcome” analysis, has approved the use of a conditional writ that returns an omitted appeal to a state court for its consideration, see Evitts v. Lucey, supra, I see no need to use the second prong of Strickland in any case where it is convenient and fair simply to return the case to the state court for further consideration, once it has been determined that counsel’s omission of a legal claim was constitutionally deficient.
. The second prong is also unnecessary in the appellate context if the omitted claim is grounded on federal law. If the habeas court determines that appellate counsel’s omission of the federal issue was below the range of prevailing professional norms, the habeas court itself is competent to adjudicate the merits of the omitted claim.
. The Appellate Division expressed no opinion on whether a state law right to effective assistance of counsel would be denied if at a preaccusation stage, interrogators were aware that a suspect was confessing on the basis of incompetent legal advice. People v. Claudio, 85 A.D.2d at 261 n. 7, 447 N.Y.S.2d at 982 n. 7.
. The two judges of the Court of Appeals who dissented in Claudio are fairly to be understood as upholding Claudio’s state law claim, even though all members of the Court of Appeals adjudicated only his federal law claim since the state law claim had not been argued. The dissenting judges expressed the view that regardless of when the federal right to counsel attaches, guaranteeing a defendant the right to the presence of counsel, the defendant has a right to the effective assistance of counsel "once counsel has been engaged.” People v. Claudio, 59 N.Y.2d 556, 565, 466 N.Y.S.2d 271, 276, 453 N.E.2d 500, 505 (1983) (Jones, J., with whom Meyer, J., joins, dissenting). Though the Supreme Court has subsequently ruled that the federal right to counsel does not attach prior to the initiation of adversary proceedings simply because an attorney-client relationship has been formed, see Moran v. Burbine, 475 U.S. 412, 430-31, 106 S.Ct. 1135, 1145-46, 89 L.Ed.2d 410 (1986), thus barring a claim of ineffective assistance of counsel, see Coleman v. Thompson, supra, the dissenting judges in Claudio would presumably have applied their view concerning a right to effective assistance of counsel whenever engaged to the state law right to counsel.
.In West, the Supreme Court ruled that the law “announced and applied [by the state intermediate appellate court] is the law of the state applicable in the same case and to the same parties in the federal court." Id., 311 U.S. at 238, 61 S.Ct. at 184. The full rigor of that ruling is not precisely applicable to a case such as this where the question is whether appellate counsel was constitutionally deficient in failing to seek reversal of the intermediate court’s decision, but the Court’s view nevertheless reenforces my belief that counsel’s omission was not incompetence.