Flores v. Demskie

11 F. Supp. 2d 299, 1998 U.S. Dist. LEXIS 7392, 1998 WL 262329

• Court: District Court, S.D. New York
• Decided: May 19, 1998
• Docket: 96 Civ. 2891(MBM)
• Status: Published

Opinion

OPINION AND ORDER

MUKASEY, District Judge.

However true it may be that, as the judicial bromide has it, hard cases make bad law, 1 it is truer still that bad law makes hard cases, this is one of them.

As set forth in greater detail below, the petitioner in this case, Rafael Flores, was convicted in Supreme Court, Bronx County, of sodomy in the first degree in violation of New York Penal Law § 130.50[1], “arising out of his deviate sexual intercourse with a six-year-old boy who lived in [his] apartment building.” People v. Flores, 84 N.Y.2d 184, 186, 615 N.Y.S.2d 662, 689 N.E.2d 19 (1994). At trial, the prosecutor failed to turn over to defense counsel until after the verdict a police officer’s memo book that contained a brief record of information arguably provided by the victim’s mother, who testified at trial. The bad law in question is a rule fashioned by the New York Court of Appeals, holding that failure by a prosecutor to turn over to a defendant before the close of evidence the prior statement of a witness called by the prosecutor at trial is automatically grounds for a new trial, with no harmless error analysis permitted if the issue is raised on direct appeal of the conviction rather than on collateral attack. Failure to deliver such material “constitutes per se error requiring that the conviction be reversed and a new trial ordered.” People v. Ranghelle, 69 N.Y.2d 56, 63, 511 N.Y.S.2d 580, 585, 503 N.E.2d 1011 (1986) (citation omitted). “The reasoning that spawned the Rosario rule [requiring such statements to be turned over to defense counsel] led us also to eschew harmless error analysis in cases arising during direct appeal in which the defendant was deprived of Rosario material at trial.” People v. Banch, 80 N.Y.2d 610, 615, 593 N.Y.S.2d 491, 494, 608 N.E.2d 1069 (1992) (citations omitted). See also People v. Young, 79 N.Y.2d 365, 370, 582 N.Y.S.2d 977, 980, 591 N.E.2d 1163 (1992). Flores’ lawyer at trial failed to take advantage of that rule, conceding instead that the entry in the memo book would have made no difference in what he did or how he questioned the mother at trial. Flores’ petition in this court, pursuant to 28 U.S.C. § 2254, argues that his lawyer’s candor deprived him ■of an automatic new trial and therefore of the effective assistance of counsel within the meaning of Strickland v. Washington, 466 U.S. 668, 104 S.Ct. 2052, 80 L.Ed.2d 674 (1984).

Because of the New York rule referred to above, this case cannot be decided based merely on a straightforward analysis of whether petitioner was deprived in any meaningful sense of a fair trial. See, e.g., Rosario v. Kuhlman, 839 F.2d 918, 925 (2d Cir.1988) (“[T]he writ would issue only where [the] petitioner can show that the error deprived [him] of a fundamentally fair trial.”) (quoting Taylor v. Curry, 708 F.2d 886, 891 (2d Cir.), cert. denied, 464 U.S. 1000, 104 S.Ct. 503, 78 L.Ed.2d 694 (1983)) (emphasis in original). Instead, what follows is a detailed review of the procedural course of this case through the state courts and before a Magistrate Judge of this court, a discussion of why I cannot affirm on the grounds set forth in the Report, 2 a brief foray into habeas corpus procedure in federal courts before enactment of the Antiterrorism and Effective Death Penalty Act of 1996, 110 Stat. 1214, (“AEDPA”), findings of fact based on an evidentiary hearing I felt the need to hold, application of the rule at issue to the facts disclosed at that hearing, and finally resolution of a state law question never examined by the courts of New York on either direct review or collateral attack of Flores’ conviction: whether the withheld material was the “duplicative equivalent” of material that was turned over, within the meaning of People v. Consolazio, 40 N.Y.2d 446, 454, 387 N.Y.S.2d 62, 66, 354 N.E.2d 801 (1976), and related cases, and therefore whether there was no violation of the rule at issue in the first place and hence no prejudice to Flores from whatever defects there may have been in the assistance provided by his counsel.

. For the reasons set forth below, that convoluted process yields the same result as would a simple harmless error analysis. Accordingly, the writ is denied and the petition is dismissed.

I.

A. Procedural History in State Court

1. Trial and Direct Appeal

On October 16, 1990, petitioner was convicted in New York State Supreme Court, Bronx County, of four counts of sodomy in the first degree and was sentenced on No■vember 9, 1990 to four concurrent terms of imprisonment of five to 15 years. (Report at 1) Petitioner’s trial attorney discovered during jury selection that the prosecution had failed to produce the memo book of a police officer who had interviewed the victim’s mother. (Id. at 2) Although the victim’s mother testified as a witness at trial, the police' officer did not. (Tr. 36) 3

Again, under New York’s Rosario rule, the prosecution must produce before trial any written or recorded statements made by a person whom the prosecution intends to call as a witness. N.Y.Crim. Proc. Law § 240.45(l)(a) (McKinney 1993) (codifying People v. Rosario, 9 N.Y.2d 286, 213 N.Y.S.2d 448, cert. denied, 368 U.S. 866, 82 S.Ct. 117, 7 L.Ed.2d 64 (1961)). The prosecution’s failure'to disclose Rosario material before the conclusion of the trial is per se reversible error. See Ranghelle, 69 N.Y.2d at 63, 511 N.Y.S.2d at 585.

Although petitioner’s trial attorney learned of the memo book entry during jury selection, he did not alert the court to its existence until after the prosecution presented its case. (Report at 2) When told that Rosario material remained undisclosed, the Court voiced exasperation with the prosecution: “The Court of Appeals has made it clear that we have to give these memo books over to the defense and here we are, the last day, thinking of charging the jury in a ease that clearly must be reversed if you don’t turn over the memo books.” (Pet. Ex. J, Trial Tr. at 167) The Court then permitted the defendant’s case to proceed and adjourned the trial until the prosecutor provided the missing memo book entry. (Report at 3) The following day, the prosecutor informed the Court that he could not obtain the memo book from the police officer. (Id.) The Court decided to charge the jury, and ordered that the “Court will have [counsel] preserve his exceptions for the record with respect to that memo book.” (Id. at 4 (quoting Trial Tr. at 238)) The Court added, “When and if the memo book becomes available, it will be turned over to counsel and if counsel feels that it’s appropriate, he will make whatever motions he deems necessary on behalf of his client even to the extent of setting aside the verdict if need be.” (Id.) After the verdict but before sentencing, the prosecutor provided defense counsel with the memo book, of which more later. (Id.)

At petitioner’s sentencing, defense counsel told the Court that he had reviewed the memo book and had concluded that “there is absolutely nothing in the memo book that would have made any difference in terms of what I did or did not do, ask or did not ask and that’s about it.” (Id. (quoting Sentencing Tr. at 2)) The Court then asked whether defense counsel wished to make a motion with respect to the verdict, and counsel moved to set aside the verdict only as against the weight of the evidence, not on the basis of the Rósano violation. (Id.)

On direct appeal, Flores argued both that the Rosario violation compelled automatic reversal and that his lawyer’s failure to argue the point at sentencing deprived him of effective assistance of counsel. In July 1993, the Appellate Division, First Department unanimously affirmed the conviction. Its reasoning, notwithstanding the per se reversible error rules described in Ranghelle and Banch, supra, was as follows:

Implicit in the rule that a guilty verdict is to be vacated in the event of non-disclosure of Rosario material is a requirement that the material might have been of some value to the defense, although the defendant is under no burden to show how (cf., People v. Jones, 70 N.Y.2d 547, 552, n. 4, 523 N.Y.S.2d 53, 517 N.E.2d 865 [1987]). When counsel candidly denies having any use for the material when it is finally disclosed, it cannot be said that the defendant has any substantive right to be vindicated, and there is no basis in law or logic to order a new trial when there is no new issue to be tried. There is no basis in the record to controvert counsel’s conclusions, and neither can counsel’s representation by deemed incompetent merely because he did not pursue a concededly frivolous substantive claim as a basis for a new trial.

People v. Flores, 194 A.D.2d 439, 599 N.Y.S.2d 255, 256 (1st Dep’t 1993). The footnote from Jones cited with the signal “cf.” is simply a discussion rejecting harmless error analysis when Rosario material has been withheld on the ground that such analysis would require “sheer speculation as to what might have occurred and matters not in the record.... ” Jones, 70 N.Y.2d at 552 n. 4, 523 N.Y.S.2d at 56 n. 4, 517 N.E.2d 865.

On July 7, 1994, the Court of Appeals affirmed, with two judges dissenting. Flores, 84 N.Y.2d at 189, 615 N.Y.S.2d at 665, 639 N.E.2d 19. The majority brushed off the underlying Rosario claim, noting that there had been “no acknowledgement that [the memo book] even technically qualified as Rosario material.” Id. 84 N.Y.2d at 186, 615 N.Y.S.2d at 663, 639 N.E.2d 19. The majority rejected the ineffective assistance of counsel claim as follows:

For all this record reveals about defense counsel’s ultimate evaluation of the disclosed item, on the scene and in the circumstances at that postverdict stage of the criminal proceeding, defense counsel could have reasonably considered that his client’s interest would best be served by no further delay in the resolution of the case against the client. Any number of other professional strategic reasons or personal considerations affecting the defendant’s fair trial interests could also have informed the attorney’s professional evaluation and choice on how and why to proceed in the way chosen. The record does emphatically inform us, even as an appellate court reviewing the matter, of the defense lawyer’s assessment of the quality and nature of the ultimately disclosed item — ‘absolutely’ useless. A plausible strategy might even have included advocacy for amelioration of sentence or some other nuanced advantage in the adversarial exchanges and context, all matters and features not knowable by trial or appellate Judges removed from the full range of choices and context exercised singularly by the trial lawyer for the defendant. In sum, Rosario was designed to insure that defense counsel, not Judges, should do the strategic viewing, weighing and exercising of the defendant’s fair trial advocacy interests in this regard.

Id., 84 N.Y.2d at 188, 615 N.Y.S.2d at 664, 639 N.E.2d 19 (citation omitted).

2. Collateral Proceedings in State Court

Unbeknownst to the Court of Appeals majority, its conjecture that defense counsel had made a strategic choice not to move for a new trial based on a Rosario violation, but instead to pursue some unspecified “nuanced advantage”, had been undermined even before those words were written. In July 1992, a year before the Appellate Division affirmed the conviction on direct appeal, Flores moved pursuant to § 440.10 of New York’s Criminal Procedure Law to set aside his conviction, based in part on ineffective assistance of counsel. Submitted with that motion was the affirmation of his trial counsel averring, under penalty of perjury, that at the time he acknowledged the uselessness of the memo book at sentencing,

I was not aware that the failure to turn over the memo book, unless waived, was per se reversible error that would have required vacatur of Mr. Flores’ conviction. I believed that to prevail on a Rosario claim I would have to show prejudice.

(Pet. Ex. L ¶ 7)

On December 4, 1992, the trial court denied the motion without a hearing. That Court acknowledged that it had virtually invited a new trial motion during colloquy about the missing memo book while the trial was ongoing, but noted that the motion it had in mind was an immediate motion pursuant to § 330.30 of New York’s Criminal Procedure Law to set aside the jury verdict, not a motion two years after the conviction pursuant to § 440.10. (Pet. Ex. G at p. 3) That Court also expressed the assumption that the disposition of the § 440.10 motion would be “incorporated in defendant’s direct appeal.” (Id.) The motion and its disposition apparently were not so incorporated, an omission that permitted the Court of Appeals to speculate about defense counsel’s quest for “nuaneed advantage” when the factual record reposing in the court below showed that defense counsel had undertaken no such quest.

On February 4, 1993, the Appellate Division, First Department, denied petitioner’s application for leave to appeal, effectively ending petitioner’s collateral attack on his conviction in the state courts.

B. Procedural History in Federal Court

1. Proceedings Before The Magistrate Judge

Petitioner then filed this petition, claiming that he was denied his right to effective assistance of counsel under the Sixth Amendment to the United States Constitution by his counsel’s failure to move to set aside the conviction based on the Rosario violation, and by his counsel’s comments concerning the uselessness of the alleged Rosario material. (Petition ¶ 22)

In his Report, Magistrate Judge Francis applied the two-part Strickland test for ineffective assistance of counsel, which requires a petitioner to show (i) that his lawyer’s conduct fell below “an objective standard of reasonableness” defined by “prevailing professional norms”, and (ii) that there is “a reasonable probability that, but for counsel’s unprofessional errors, the result of the proceeding would have been different.” Strickland, 466 U.S. at 688, 694, 104 S.Ct. 2052. Under Strickland, “the performance inquiry must be whether counsel’s assistance was reasonable considering all the circumstances” and “the defendant must overcome the presumption that, under the circumstances, the challenged actions ‘might be considered sound trial strategy.’ ” 466 U.S. at 688-89, 104 S.Ct. 2052 (quoting Michel v. Louisiana, 350 U.S. 91, 101, 76 S.Ct. 158, 100 L.Ed. 83 (1955)). Without holding an evidentiary hearing to determine whether the withheld memo book was in fact Rosario material, the Magistrate Judge found that although defense counsel’s performance was deficient, petitioner had not established the prejudice prong of the Strickland test.

The Magistrate Judge found that counsel’s failure to move for a new trial based on per se reversible error was objectively unreasonable. He found that the statements in counsel’s affidavit that he was unaware of the per se rule and thought he would have to make a showing of prejudice, show that his decision not to press the Rosario issue did not arise from “either a reasonable professional judgment or a reasoned strategic choice, but rather a complete misunderstanding of prevailing law.” (Report at 12).

However, he found that petitioner had failed to meet Strickland’s prejudice prong. He reasoned that the prejudice inquiry differs for trial and appellate counsel. According to the Magistrate Judge, the test to be applied to determine prejudice resulting from trial counsel’s performance is “ ‘whether counsel’s deficient performance renders the result of the trial unreliable or the proceeding fundamentally unfair.’” (Report at 13) (quoting Lockhart v. Fretwell, 506 U.S. 364, 372, 113 S.Ct. 838, 122 L.Ed.2d 180 (1993)) On the other hand, to establish prejudice resulting from appellate counsel performance, a petitioner must show “ ‘that there was a reasonable probability that [his] claim would have been successful before the [state’s highest court].’ ” (Id.) (quoting Mayo v. Henderson, 13 F.3d 528, 584 (2d Cir.), cert. denied, 513 U.S. 820, 115 S.Ct. 81, 130 L.Ed.2d 35 (1994)) The Magistrate Judge applied the trial counsel standard and found that petitioner had failed to demonstrate that his attorney’s deficient performance rendered his trial unreliable or unfair. (Id. at 14-15).

2. Proceedings Before This Court

(i) Objections to the Report

Petitioner objects to the Magistrate Judge’s decision relating to the prejudice prong, and respondent to the Magistrate Judge’s disposition of the performance prong. Pursuant to Fed.R.Civ.P. 72(b), a district court may adopt those parts of a magistrate judge’s report and recommendation to which no specific objection is made, provided those parts are not clearly erroneous. Thomas v. Arn, 474 U.S. 140, 149, 106 S.Ct. 466, 88 L.Ed.2d 435 (1985). However, the district court must make a de novo determination regarding those parts of the report to which any party objects. See id. Because the parties object to the Magistrate Judge’s findings on both prongs of the Strickland test, a de novo review as to both is necessary.

(ii) The Report Considered

I cannot affirm for the reasons contained in the Report. To begin with, I disagree with the Magistrate Judge’s conclusion that petitioner’s attorney was acting as trial counsel, rather than as appellate counsel, at the sentencing proceeding. Counsel made the alleged error after the completion of the trial, and the error concerned the failure to make a motion to set aside the verdict on a particular ground. At that stage, as petitioner argues, defense counsel was acting in the role of appellate counsel, i.e., reflecting on the conduct of the trial and raising errors of state or federal law. Counsel’s failure to make this motion amounted to a virtual waiver of the issue for appeal.

In any event, I disagree with the Magistrate Judge’s differentiation between the prejudice inquiry for trial and appellate counsel. There is no basis in Second Circuit case law for distinguishing between trial and appellate counsel prejudice. See Bunkley v. Meachum, 68 F.3d 1518, 1521, 1523 (2d Cir.1995) (applying undifferentiated Strickland standard to conduct of appellate counsel); Mayo, 13 F.3d at 533 (same); Claudio v. Scully, 982 F.2d 798, 803 (2d Cir.1992) (“Although Strickland addressed the constitutional standard for ineffective assistance of counsel in the trial counsel context, our Circuit has adopted the Strickland two-prong test in assessing the effectiveness of appellate counsel.”), cert. denied, 508 U.S. 912, 113 S.Ct. 2347, 124 L.Ed.2d 256 (1993).

Further, the Magistrate Judge relied too heavily upon Fretwell to make this distinction between trial and appellate counsel and to require a showing of unreliability in the trial or unfairness in the proceeding for every error by counsel. In Fretwell, the petitioner claimed that his attorney was ineffective because his counsel failed to object at the sentencing proceeding to an error that at the time mandated reversal of the death penalty under Eighth Circuit precedent. However, in the interim, the Eighth Circuit had overruled its earlier decision, and the alleged defect was no longer considered a constitutional violation necessitating reversal. Fretwell, 506 U.S. at 367-68, 113 S.Ct. 838. The Court stated that “an analysis focusing solely on mere outcome determination, without attention to whether the result of the proceeding was fundamentally unfair or unreliable, is defective.” Id. at 369, 113 S.Ct. 838. The Court found that the sentencing proceeding was not unfair or unreliable because “counsel’s error would have deprived respondent of the chance to have the state court make an error in his favor.” Id. at 371, 113 S.Ct. 838 (quotations and citations omitted). The Court added, “Unreliability or unfairness does not result if the ineffectiveness of counsel does not deprive the defendant of any substantive or procedural right to which the law entitles him.” Id. at 372, 113 S.Ct. 838. Thus, in Fretwell, the Court declined to find prejudice when, although counsel’s objection would haye altered the outcome of the proceeding, that change in outcome would have been based on an incorrect interpretation of the law which had since been corrected.

Fretwell’s scope is limited. The Strickland test—whether there is “a reasonable probability that, but for counsel’s unprofessional errors, the result of the proceeding would have been different,” Strickland, 466 U.S. at 694, 104 S.Ct. 2052—was not overruled by Fretwell. Rather, as Justice O’Con-nor stated in her concurring opinion, the Court’s holding was narrow: “Specifically, today we hold that the court making the prejudice determination may not consider the effect of an objection it knows to be wholly meritless under current governing law, even if the objection might have been considered meritorious at the time of its omission.” Fretwell, 506 U.S. at 374, 113 S.Ct. 838 (O’Connor, J., concurring).

The Second Circuit has similarly recognized Fretwell’s narrow holding. In Bunk-ley, the Second Circuit faced the problem in Fretwell—i.e., the state courts had misinterpreted federal precedent, and thus although the outcome in state court would have been different had counsel raised an issue on appeal, a proper interpretation of federal law would not have altered the result. Bunkley, 68 F.3d at 1520-21. Bunkley, in discussing Fretwell, noted that when a petitioner claims ineffective assistance of appellate counsel, “where the underlying claim involves only state law, looking to the probable outcome of the direct appeal in state court will be adequate under federal constitutional doctrine to preserve confidence in the outcome and protect the defendant from prejudice.” Id. at 1522-23. 4 Thus, Fretwell does not dictate that a Court always inquire into the reliability and fairness of the proceeding in addition to whether the outcome would have been different; rather, its fairness and reliability inquiry applies only in limited circum stances—i.e., when the state or lower court creates a rule favorable to the defendant based on an erroneous interpretation of federal law. See also Grady v. Artuz, 931 F.Supp. 1048, 1062 (S.D.N.Y.1996); see also Borla v. Keane, 99 F.3d 492, 497 (2d Cir.) (applying outcome-determination test to trial counsel performance), clarified on reh’g on other grounds, 90 F.3d 36 (1996), cert. denied, — U.S. -, 117 S.Ct. 2508, 138 L.Ed.2d 1012 (1997).

Here, the additional Fretwell inquiry is not required. There is no incorrect state interpretation of federal law at issue. However, in order to determine whether there would have been merit to the argument defense counsel failed to make, it is necessary to determine whether the undisclosed material here justified making it. If it did, then defense counsel failed to raise an issue of state law which appears well settled and which the Court of Appeals has recently recognized. See People v. Machado, 90 N.Y.2d 187, 191, 659 N.Y.S.2d 242, 244-45, 681 N.E.2d 409 (1997). The outcome-determination test applies.

Moreover, even if Fretwell’s fairness and reliability inquiry did apply, it would not dictate a different result from the outcome determination test. That Fretwell test requires that the Court determine whether the proceeding was unfair, i.e., whether petitioner was deprived of any substantive or procedural right. Here, petitioner argues that he was deprived at least of a new trial because his lawyer was not aware that he could have moved for one without showing prejudice. A defendant with competent counsel, petitioner argues, would have received a new trial, while petitioner did not. This, if he is correct, was unfair.

Finally, I cannot accept respondent’s argument that the Fretwell test applies to every ineffective assistance of counsel claim, because that would mean that ineffective assistance of appellate counsel could never be found unless the trial was unreliable or unfair. However, this is clearly not the law. See Mayo, 13 F.3d at 534 (“To establish prejudice in the appellate context, a petitioner must demonstrate that [his] claim would have been successful before the [state’s highest court].”) (quotations and citations omitted; alterations in original). Rather, prejudice is demonstrated when the argument that appellate counsel failed to raise on appeal would have been successful when properly analyzed under controlling law.

(iii) The Evidentiary Hearing and the Rosario Issue

Among respondent’s arguments opposing the Magistrate Judge’s finding of ineffective assistance was that there had been no showing that the police memo- book in question contained Rosario material. Absent such a showing, respondent argued, counsel could not have been ineffective in failing to move for a new trial based on non-production of Rosario material.

Although one would assume that the trial court to which such a motion was addressed would have held a hearing to resolve that issue, see Young, 79 N.Y.2d at 372, 582 N.Y.S.2d at 981, 591 N.E.2d 1163 (remanding to trial court for hearing to determine whether withheld report contains Rosario material), respondent opposed such a hearing in this court. Respondent argued both that the revisions to 28 U.S.C. § 2254 wrought by the AEDPA bar such a hearing, and that prior law requires the same result because petitioner’s failure to develop the record on that issue in state court cannot be excused by cause external to himself and resulting prejudice. Subsection (e)(2) of § 2254 does bar evidentiary hearings in federal court to develop the factual basis for a claim unless that claim arises from either newly discovered evidence or a new rule of constitutional law that is retroactive, and the facts underlying that claim would show by clear and convincing evidence that no reasonable jury would have convicted the petitioner. Neither is true here, but this petition was filed before the effective date of that section, and the statute is not retroactive. See Lindh v. Murphy, 521 U.S. 320, -, 117 S.Ct. 2059, 2068, 138 L.Ed.2d 481 (1997); Boria, 90 F.3d at 38.

Moreover, under the statute before it was revised by the AEDPA a district court had discretion to grant a hearing in all eases, and was required to hold an evidentiary hearing when the failure to develop the record in state court was excused by cause and prejudice. See Pagan v. Keane, 984 F.2d 61, 64 (2d Cir.1993). Factors relevant to the exercise of that discretion include: the existence of a factual dispute, the strength of the proffered evidence, the thoroughness of prior proceedings, and the nature of the state court determination. Id. In this case, there was a factual dispute as to whether the memo book constituted Rosario material; the proffered evidence, consisting of the memo book entries and portions of the testimony of the victim’s mother, suggested she might have been the source of some of those entries, although respondent argued that so might the child’s school guidance counsellor, who did not testify at trial; prior proceedings had not determined this issue; and the state courts had made no findings on this issue. Moreover, even under respondent’s view of the law, it is at least arguable that trial counsel’s decision to forgo the Rosario issue, based not on an informed and strategic choice but rather on ignorance, was a cause external to Flores, that prejudice could be shown to have resulted if it were shown that the withheld record contained Rosario material, and that the lost opportunity even to try to make that showing was itself prejudice. In these circumstances, I elected to hold an evidentiary hearing.

Two witnesses testified at the hearing. One was former police officer Karen Pabon, the author of the memo book entries at issue (Pét.Ex. 2) and the author also of a police report that was turned over as Rosario material at trial. (Pet.Ex. 3) Not surprisingly, she had only a vague recollection of the events, which had occurred in 1989 and had resulted in a trial in which she was not called as a witness. (Tr. 15, 29, 30, 36, 38, 57) However, she did testify at the hearing, based principally on her memo book entries and the police report she had prepared, that while she was on duty on January 30, 1989, she received a call at about 11:40 a.m. reporting a sexual assault, and responded to P.S. 211 in the Bronx. There, she spoke to the victim’s mother. Because the mother spoke only Spanish and the officer did not, the two conversed through the school guidance coun-sellor, who acted as interpreter. (Tr. 29) The idctim was present, but the officer did not recall speaking with him. (Tr. 38)

The" principal issue in dispute at the hearing was the source of entries in Pabon’s memo book and her report. Apparently hoping to fit this case within a line of state authority-holding that statements reflecting merely a synopsis of events reported by several witnesses may not constitute Rosario material as to any particular witness, see People v. Scott, 216 A.D.2d 592, 595, 628 N.Y.S.2d 965, 968 (2nd Dep’t 1995), aff'd, 88 N.Y.2d 888, 644 N.Y.S.2d 913, 667 N.E.2d 923 (1996), respondent’s counsel pressed the officer to characterize her entries as reflecting such a synopsis, at times successfully. (See Tr. 43-48) However, I attach little weight to those characterizations because the content of the memo book discloses entries containing statements that could have come only from the mother direct to Pabon, based on Pabon’s testimony that she spoke to the mother and based also on the police report she wrote, which listed the mother as the reporter of the information it contained. That information was, as set forth more fully below, equivalent in all comparable respects with the corresponding information in the memo book.

Respondent called Lieutenant George Stamp, who at the time of the events in question had been a sergeant and who prepared an unusual incident report 5 on January 30, 1989 based on information conveyed to him by Pabon and another officer. (Pet. Ex. A) That report reflects that Pabon apparently attributed to the guidance counsel-lor, Hilda Beyares, the information that was attributed to the mother in the police report. From Stamp’s unusual incident report, respondent apparently would have me conclude that it was the guidance counsellor and not the mother who provided the information in the memo book, and, inasmuch as the guidance counsellor did not testify at trial, that the memo book entries are not Rosario material with respect to the mother, who did testify. I cannot so conclude. Based on Pabon’s testimony at the hearing, and a comparison of the memo book entries with the police report she wrote, I find that portions of the memo book that I have specified below do reflect statements from the mother, who testified at trial. To the extent Stamp’s unusual incident report suggests that information came from the guidance counsellor, I find that it simply reflects either Stamp’s misconception, or a misstatement by Pabon or her colleague. Therefore, the memo book, considered in isolation, does contain Rosario material with respect to the victim’s mother. Set forth below are all of the entries from the officer’s memo book that relate to the incident, line by line, with spelling as in the original and those entries that I find attributable to the mother set forth in italics:

1140-Sexual Assault P.S.211-1155-1 female & 1 juvenile to Jacobi Hosp-1210-At Jacobi Hosp-Examination done by Dr. Cohen — Sex Crimes Notified Castellano-Det Gannon is Catching-61 # 1145 Aided # 417 Borolog # 317 Sex Crimes Case # 111 Victim [name of victim 6 7 was sodomized by neighbor while he went to play their with perps children - Victim notified mother Who in turn notified Guidance counsler Hilda Beyares. Mother-Carmen Santiago DOB / / present at school & at Hospital 1445 EnRt to 48 to Sex Crimes Unit Det Gannon - Perp Rafael flores 1997 Hughes Ave Apt 1L approx 32 M/H- 1500-At 48 pet - Brady by PO Gleason - 1505 10-61 Sgt Stamp doing an unusual D I Luisi Notified PBBS-Schrepel Notified

(Pet.Ex. 2) The reason for finding that these entries must reflect statements by the victim’s mother should be obvious: As noted, the officer said she did not speak with the victim, and the guidance counsellor acted only as a translator; by a process of elimination, only the mother could have conveyed the italicized information.

As was apparent from the hearing, however, the memo book does not stand in isolation. More particularly, the entries in the memo book and those in Pabon’s report, to the extent attributable to the victim’s mother, were based on the same conversation. (Tr. 29-30) Nobody disputes that Pabon’s report was turned over as Rosario material at trial. Set forth in the table below are the entries from the memo book, italicized above, that I have found attributable to the mother, and therefore, also Rosario material, alongside those entries from Pabon’s report that are also attributable to the mother. Again, spelling is as it appears in the original documents.

Memo Book

Victim [name of victim] was sodomized by neighbor while he went to play their with perps children-victim notified mother Mother-Carmen Santiago Perp Rafael flores 1997 Hughes Ave Apt 1L approx 32 M/H-

Report 7

[Offense(s) ] Sexual Abuse (Sodomy) [Victim] [name of victim] [Actions of Victim Prior to ... Sex Crimes] playing at neighbors apt. [Reporter/Witness No. 1] Santiago, Carmen [Relationship] mother [Perpetrator] Flores, Raphael 1997 Hughes Ave., Apt. 1L Bx, N.Y. [Sex] M [race] Hisp [date of Birth] unknown [Age] 32 [Height] 5Ft. 6In. [Weight] 185 [Eye Color] unkn [Hair Color] blk [Hair Length] short [Facial Hair] none [Accent] hisp [Clothing Description] Unknown [Reconstruct Occurrence] At Vp/o reporter states for victim that while going to a neighbor’s house to play the above described perp forced victim to have anal sex with him (penis to anus). Victim states that this has been an on going occurrence for the past years.

Following the hearing, I asked the parties for submissions on the question of whether the statements in the report attributable to the mother were the “duplicative equivalent” of those so attributable in the memo book, within the meaning of People v. Consolazio, 40 N.Y.2d at 454, 387 N.Y.S.2d at 66, 354 N.E.2d 801. Consolazio held that there can be no Rosario violation in failing to turn over statements that are “nothing more than the duplicative equivalents of statements previously turned over to the defense—the only difference being as to the .particular form in which such statements are recorded.” Id. Neither side had briefed or argued that issue until then. Petitioner responded, inter alia, with a long list of differences between the entries in Pabon’s memo book and those on the report (Jan. 23, 1998 Letter of Jessica S. Henry, Esq. (“Henry Letter”), at 4-6), a list that is irrelevant insofar as. it consists, as it largely does, of entries that obviously do not reflect statements attributable to the victim’s mother and therefore are not Rosario material as to her. (See, e.g., id. at 4, items a-i) However, petitioner argues that because the memo book and the report each contains material that is not found in the other, and omits material that is found in the other, neither can be considered the duplicative of the other. (Id. at 3)

Respondent’s submission argues principally that because petitioner had failed to preserve the argument that the memo book contained Rosario material, “this factual issue cannot be considered by this Court because it is unexhausted,” such lack of exhaustion being “a direct result of petitioner’s failure to present the state court with ... a preserved claim of Rosario error.” (Jan. 27, 1998 letter of John F. Carroll, Esq. (“Carroll Letter”), at 2) When one considers that the only point being disputed in this petition is whether defense counsel at trial was ineffective for failing to preserve the Rosario issue, and that the only way to avoid that finding that was under discussion at the time I asked counsel for submissions was the possibility that the memo book might contain only the duplicative equivalent of Rosario material, respondent’s insistence that the issue had not been preserved was unresponsive, and less than helpful to his own position. Respondent’s only relevant answer to the question I posed was to quote from a segment of Pabon’s report, and then to assert, although not to demonstrate, that “the complaint report is virtually identical to the memo book entry.” (Carroll Letter at 2) Respondent also pointed out that the failure to argue the duplicate equivalency issue in the courts of New York, and' the consequent failure of those courts to address the issue, does not bar this court from addressing it. See Pinkney v. Keane, 920 F.2d 1090, 1094 (2d Cir.1990), ce rt. denied, 501 U.S. 1217, 111 S.Ct. 2824, 115 L.Ed.2d 995 (1991). (Id. at 1 n. 1) There is no unfairness to petitioner in reaching that issue here. Respondent could not have been expected to argue in state court that the memo book was the duplicate equivalent of the police report if petitioner did not move for a new trial based on the failure to produce the memo book. It is the absence of such a motion that is the basis for petitioner’s argument here, and that occasions an examination of the duplicate equivalency issue.

II.

Prior to Rosario, the trial courts in New York State determined which documents in the prosecution’s possession were relevant to the defense and ordered production accordingly. See People v. Machado, 90 N.Y.2d 187, 191, 659 N.Y.S.2d 242, 244, 681 N.E.2d 409 (1997). Under the Rosario rule, later codified in a statute, the prosecutor must “make available to the defendant ... [a]ny written or recorded statement ... made by a person whom the prosecutor intends to call as a witness at trial, and which relates to the subject matter of the witness’s testimony.” N.Y.Crim. Proc. Law § 240.45(1)(a). The rationale for the change in approach, as articulated in Rosario, was that as between a trial judge and defense counsel, “the latter is in a far better position to appraise the value of a witness’ pretrial statements for impeachment purposes,” and therefore the defense should have access to all such statements before trial. Rosario, 9 N.Y.2d at 290, 213 N.Y.S.2d at 451, 173 N.E.2d 881.

That rationale was later extended still further by the Court of Appeals. Defense counsel’s “far better position to appraise the value of a witness’ pretrial statements for impeachment purposes” later came to mean that judges could have no say at all in that appraisal. In Consolazio, 40 N.Y.2d at 454, 387 N.Y.S.2d at 66, 354 N.E.2d 801, the Court of Appeals announced, although it did not apply, a rule that eschewed harmless error in direct review of convictions involving alleged Rosario violations. As later articulated by the Court of Appeals, the rationale for that per se rule was as follows:

The essence of the Rosario requirement, that the prosecutor supply all of a witness’ statement or statements relating to his testimony, is that a judge’s impartial determination as to what portions may be useful to the defense, is no substitute for the single-minded devotion of counsel for the accused. Appellate judges have no advantage over trial judges in making such a determination.

People v. Perez, 65 N.Y.2d 154, 160, 490 N.Y.S.2d 747, 750-51, 480 N.E.2d 361 (1985). In a later case, the Court pronounced it flatly “impossible” to engage in harmless error analysis of Rosario violations because to do so would involve “sheer speculation” on such subjects as “the questions counsel might have asked or avoided asking, how the witness might have answered, and how such differences in the testimony might have affected the decision of the jury.” Jones, 70 N.Y.2d at 551 and 552 n. 4, 523 N.Y.S.2d at 56 and n. 4, 517 N.E.2d 865. “It is this inability of a reviewing court, to ascertain how defense counsel might have employed the withheld material, that distinguishes cases involving Rosario violations from cases involving other errors where the application of harmless error analysis is appropriate.” Jones, 70 N.Y.2d at 552, 523 N.Y.S.2d at 56, 517 N.E.2d 865.

Actually, this purported inability to determine how defense counsel might have employed withheld information is not unique to cases involving Rosario violations, yet the Court of Appeals has mandated harmless error review in similar cases. See People v. Steadman, 82 N.Y.2d 1, 8-9, 603 N.Y.S.2d 382, 385, 623 N.E.2d 509 (1993) (requiring harmless error analysis in weighing claim of withheld impeachment material). Even further, this purported inability in Rosario cases is found to afflict trial and appellate judges only when a conviction is still on direct review, the New York State Legislature having failed to prescribe a standard of review when it codified the Rosario rule in N.Y.Crim. Proc. Law § 240.45(l)(a). However, when a conviction is collaterally attacked by motion, the statute requires a petitioner to show prejudice. N.Y.Crim. Proc.- Law § 440.10(l)(f) (McKinney 1993). Recognizing that it was not free simply to disregard that requirement, the Court of Appeals simply ratified two standards of review of alleged Rosario violations: one, a judge-made rule, bars judges engaged in direct review of a conviction from determining whether a withholding of Rosario material has resulted in prejudice to the defendant because, as stated in Perez and Jones, judges are unsuited to the task and therefore they can not, should not and will not do it; the second, a rule made by the legislature, requires judges to make that determination and therefore, although judges have pronounced themselves unsuited and unable to do it, they nonetheless do it. See People v. Jackson, 78 N.Y.2d 638, 647, 578 N.Y.S.2d 483, 489, 585 N.E.2d 795 (1991). Different standards of review before and after appeals had been exhausted conceivably could be rationalized based on society’s lesser interest in the finality of convictions before such exhaustion, but harmless error review is applied to a motion under N.Y.Crim. Proc. Law § 440.10 even when that motion is made before the defendant’s appeal has been concluded. See Machado, 90 N.Y.2d at 193, 659 N.Y.S.2d at 245, 681 N.E.2d 409.

The per se reversal rule has also been rationalized in part as a judicially created incentive to prosecutors to disclose Rosario material in timely fashion, see Consolazio, 40 N.Y.2d at 455, 387 N.Y.S.2d at 67, 354 N.E.2d 801 (“Reflection thus suggests that once it is determined that the writings sought by the defendant come within the Rosario rule, the better practice would be to direct a turnover forthwith.”). It bears mention that not only did that incentive fail here, but there was also displayed an incentive of a different sort, once the memo book did emerge, to craft a former officer’s testimony in order to make it appear that the memo book did not contain Rosario material, but rather contained a synopsis of events not attributable to any particular witness. (See p. 308, supra) Moreover, on another occasion, the Court of Appeals noted that in such cases, the diligence of the prosecutor or absence of it does not matter at all, any more than the significance of the material or lack of it; this observation was made even while presenting the rationale of Rosario as one arising from fairness. Jones, 70 N.Y.2d at 553, 523 N.Y.S.2d at 57, 517 N.E.2d 865 (“The focus of Rosario is on fairness to defendant — not on the conduct or motives of the prosecutor. If Rosario material is denied the defendant, he has been deprived of what he should have. It matters not that the denial may have been inadvertent or immaterial.”). I have been unable to find the case that explains in what sense reversing a conviction for inadvertent failure to disclose immaterial information promotes “fairness to defendant.”

The anomalies arising from the per se reversal rule, such as raising the consequences of nonconstitutional Rosario violations higher than those of constitutional violations, which may be subject to harmless error analysis, have been recognized even by members of the Court that created the rule, see Jones, 70 N.Y.2d at 553-57, 523 N.Y.S.2d at 57-60, 517 N.E.2d 865 (Bellacosa, J., concurring), but the force of the rule remains undiminished. See Banch, 80 N.Y.2d at 615, 593 N.Y.S.2d at 494, 608 N.E.2d 1069. Indeed, the rule has resulted in at least one reversal in federal court under § 2254 based on a theory identical to the one petitioner argues here: that counsel who failed to press a viable Rosario claim so as to seek automatic reversal of a conviction was, for that failure, ineffective under Strickland. Mayo, 13 F.3d at 537.

However, as noted, the New York Court of Appeals recognized in Consolazio an .exception to the per se reversal rule for cases in which the Rosario material that is not dis closed in timely fashion is the “duplicative equivalent” of Rosario material that is so disclosed. In Consolazio, a fraud ease, the prosecutor turned over the grand jury testimony of complaining witnesses who testified at trial, but failed to turn over questionnaires filled out by an investigator who had spoken with the witnesses and wrote down their answers to questions about how they had come to meet the defendant and to invest money with him. Consolazio, 40 N.Y.2d at 452, 387 N.Y.S.2d at 65-66, 354 N.E.2d 801. The Court of Appeals found that the undisclosed questionnaires, referred to as “worksheets,” were Rosario material as to the witnesses whose statements were recorded therein, 40 N.Y.2d at 453, 387 N.Y.S.2d at 65, 354 N.E.2d 801, but nonetheless concluded that reversal was not warranted:

Our examination of the Grand Jury' testimony of the various prosecution witnesses (which testimony was turned over by the prosecutor to the defense) reveals that the witnesses’ statements contained in the worksheets were the same as the statements made by such witnesses before the Grand Jury. The worksheets in this instance were nothing more than duplicative equivalents of statements previously turned over to the defense — the only difference being as to the particular form in which such statements were recorded. In this circumstance it was not error to fail to turn over worksheets which would have been cumulative only.

40 N.Y.2d at 454, 387 N.Y.S.2d at 66, 354 N.E.2d 801. Although it seems counter-intuitive that the statements on the questionnaires were literally “the same” as the grand jury testimony, and examination of the appendix to the brief submitted by Consolazio to the Court of Appeals tends to confirm what intuition suggests, compare Consolazio App.2018 (quoting grand jury testimony that was disclosed) with Consolazio App.2019-20 (quoting questionnaire responses of the same witness), the Court of Appeals nonetheless concluded that they were sufficiently similar not to warrant reversal.

In Ranghelle, however, an opinion that treated appeals in two cases, the Court of Appeals appears to have applied the duplica-tive equivalent exception more rigidly. In the first of the two cases before the Court, the only item of Rosario material was undisclosed. When compared with the witness’ trial testimony, the undisclosed Rosario material “was consistent with [the witness’ trial] testimony, [but] it lacked essential details to which she later testified at trial, omissions which might have constituted important material for cross-examination.” Ranghelle, 69 N.Y.2d at 64, 511 N.Y.S.2d at 585, 503 N.E.2d 1011. The second case decided in the Ranghelle opinion is closer to the case at bar in that one item of Rosario material — a police report — had been turned over, while another — officers’ memo books — had not. However, although both items of Rosario material apparently recorded the same statements of the complaining witness, it was not as to the complaining witness that the Rosario issue was raised, but rather as to the officers, who testified at trial. As to them, the memo book entries constituted statements at a different time from the statements recorded in their report. The Court ruled that when there were unspecified “inconsistencies” between the two records, albeit “minor,” the memo books could not be considered “duplicative equivalents” of the police report within the meaning of Consolazio and reversal was warranted. 69 N.Y.2d at 65, 511 N.Y.S.2d at 586, 503 N.E.2d 1011. In its most recent ruling on the duplicative equivalent issue, the Court has emphasized the absolute necessity of comparing undisclosed and disclosed records, when more than one record is at issue. See People v. Joseph, 86 N.Y.2d 565, 569, 635 N.Y.S.2d 128, 125, 658 N.E.2d 996 (1995), where the Court held that if the undisclosed record is missing or destroyed, and no comparison is possible, no argument of equivalency can be accepted.

In Jones, as in the first of the two cases decided in Ranghelle, only undisclosed Rosario material was at issue. The concurrence argued, as the prosecution had in Ranghelle, that that undisclosed material was the “duplicative equivalent” of the witness’ trial testimony. Jones, 70 N.Y.2d at 553-57, 523 N.Y.S.2d at 57-60, 517 N.E.2d 865 (Bella-cosa, J., concurring). As in Ranghelle, the argument was rejected. The undisclosed material showed, as the witness’ trial testi mony had not, that the witness had been arrested on a weapons charge, that he had been fired from his job, and that he had “actually sold 2]£ to 5 times more than the quantity of drugs which he claimed in his trial testimony to have sold.” 70 N.Y.2d at 550, 523 N.Y.S.2d at 54-55, 517 N.E.2d 865.

Finally, in Young, a murder case, the undisclosed Rosario information included a description of potentially significant marks on the victim’s body. The Court treated the discrepancies between statements as a bar to application of the “duplicative equivalent” rule, and used language that petitioner emphasizes, at least in part, in his submission here:

Indeed, a statement that is consistent with other disclosed material but omits details or facts cannot be considered the “duplica-tive equivalent” of the disclosed material, since omissions often furnish important subjects for cross-examination.

Young, 79 N.Y.2d at 370, 582 N.Y.S.2d at 980, 591 N.E.2d 1163 (emphasis in Henry Letter at 2).

III.

The key to deciding this case, in my view, lies in the unemphasized final clause of the sentence from Young quoted immediately above, which explains the need for a close comparison of disclosed to withheld statements as follows: “omissions often furnish important subjects for cross-examination.” So they do, but only when the omissions at issue appear in statements by the witness at different times. If the omissions at issue appear merely in different records of the same statement by the witness, then any omissions in an undisclosed record cannot conceivably be the subject of cross-examination of the witness because those omissions arise merely ftwn variations in what the recorder of the statement wrote in different places, not from variations in what the witness said on different occasions, the witness by hypothesis having spoken only once.

In this case, according to Pabon’s testimony, we are dealing with two documents, the memo book and the police report, that record the same conversation with the witness. (Tr. 29-30) Any omissions in the memo book, which was not turned over, as compared with the police report, which was, could not have been the subject of cross-examination of the witness, the victim’s mother, because she did not write either. Therefore, omissions in the memo book as compared to the police report, if such there be, are irrelevant. The only conceivably relevant inquiry is whether there is information in the memo book attributable to the mother that is not included in the police report, because such information would have been unavailable to counsel for cross-examination purposes.

That inquiry, as demonstrated by the chart on page 309 above, shows that the memo book, like the police report, reflects the name of the victim; the crime committed against him; the name of the mother; and the name, address, approximate age and ethnicity of the perpetrator. The memo book says “victim notified mother”; the police report does not say that in so many words, but it reflects that the mother was the person who reported the crime and thus, obviously, that her child told her about it. The memo book states that the victim “was sodomized by neighbor while he went to play their [sic ] 8 with perps children.” The police report states that the crime occurred when the victim was “playing at neighbors apt.” and “while going to a neighbor’s house to play.” In this regard, the only information that appears in the memo book that does not appear explicitly in the police report is that the victim had gone to play at the perpetrator’s apartment “with perps children.” That fact, however, is utterly unremarkable; it simply completes the thought that the victim had gone to the perpetrator’s apartment to play with those whom children most commonly play with when they visit the apartments of neighbors — the neighbors’ children. The phrase, “with perps children” conveys nothing that is not implicit by common experience in the disclosure that the victim had gone to the neighbor’s house to play. Certainly, that phrase cannot convey any inconsistency in prior statements by the mother, who is the only witness at issue, because, again, we are dealing with two items of Rosario material that record a single statement on a single occasion, not two statements on separate occasions. Moreover, although Ranghelle and its progeny foreclose any inquiry into the usefulness of Rosario material, I note that there has been no claim that the phrase was inconsistent with the witness’ trial testimony, or otherwise could have provided the basis for any inquiry on cross-examination.

No New York case cited by counsel, or that I have been able to locate, deals with a situation where, as here, one statement was recorded in two different places, and the failure to turn over one of those records necessitated a new trial because of a finding that such failure constituted a Rosario violation with respect to the witness who made the statement. 9 Accordingly, consistent with controlling New York authority and, in particular, with the exception to the per se reversal rule recognized in Consolazio and its progeny, I find that in this case the undisclosed memo book was the duplicative equivalent of the disclosed police report. Therefore, there is no basis for finding that a new trial would have been granted here, even if Flores’ trial counsel had moved for one. Nor is there any basis, therefore, to find that Flores suffered prejudice from his lawyer’s concession that the memo book entries would have been useless in cross-examining the victim’s mother. See Strickland, 466 U.S. at 694, 104 S.Ct. 2052.

For the above reasons, the writ is denied and the petition is dismissed.

SO ORDERED.

1 . Probably the most frequently cited invocation of that hypothesis is Justice Holmes’ statement, dissenting in the Northern Securities case, that “Great cases, like hard cases, make bad law.” Northern Securities Co. v. United States, 193 U.S. 197, 364, 24 S.Ct. 436, 48 L.Ed. 679 (1904). He took it as well established, even in 1904, that hard cases make bad law, and went on to assert that the same is true of great cases because great cases are so called due to their tendency to excite "overwhelming interest which appeals to the feelings and distorts the judgment. These immediate interests exercise a kind of hydraulic pressure which makes what previously was clear seem doubtful, and before which even well settled principles of law will bend.” Id, I have also heard it said that easy cases make bad law because their apparent simplicity invites carelessness of thought and word, which results in rules ill suited to more complex cases.

2 . References to the "Report” are to the May 13, 1997 Report and Recommendation of Magistrate Judge James C. Francis IV, to whom this case was initially referred and who recommended denial of the petition. I reach the same result recommended by the Magistrate Judge, although for different reasons. Because both petitioner and respondent object to the Report, I have relied on it only insofar as it is either not controverted or agrees with my own reading of the record, which I have reviewed de novo.

3 . "Tr." refers to the transcript of an evidentiary hearing held on December 12, 1997.

4 . Significantly, although the Magistrate Judge suggested that Fretwell was applicable only to ineffective assistance of trial counsel claims. Bunkley applied it to an ineffective assistance of appellate counsel claim. Bunkley, 68 F.3d at 1523.

5 . Such a report, as its name suggests, describes . an incident, usually a crime, that is beyond the usual mayhem the police encounter in the precinct in question. (Tr. 58)

6 . The victim’s name appears in the record, but has been omitted from this opinion.

7 . Bracketed words are printed headings on the form, except for brackets indicating deletion of the victim's name; words outside brackets are typed entries following or beneath those headings.

8 . This is an obvious misspelling of “there”, meaning at the neighbor’s apartment.

9 . I have found only one case involving such a failure to turn over a statement recorded in two different places where that argument was even made, see People v. Gaskins, 171 A.D.2d 272, 575 N.Y.S.2d 564 (2nd Dep't 1991), and that case was decided on statutory grounds and not based on whether the undisclosed statement was or was not the duplicative equivalent of the disclosed statement. Gaskins involved a child victim of sexual abuse whose testimony was videotaped for presentation to the grand jury pursuant to N.Y.Crim. Proc. Law § 190.32 (McKinney 1993). The transcript was turned over to defense counsel but the tape itself was not. However, N.Y.Crim. Proc. Law § 240.45, the statutory embodiment of the Rosario rule, requires that both the videotaped testimony and the transcript be disclosed, and on that statutory basis the appellate court directed a new trial. Although Flores suggests that People v. Gallman, 240 A.D.2d 512, 657 N.Y.S.2d 781 (2d Dep't 1997) involved the same statement recorded in two different places, that decision, a summary memorandum, does not so state, but does report the prosecutor's concession that a duplicative equivalent analysis could not be made. Id., 657 N.Y.S.2d at 782. There is no such concession here.