(dissenting in People v Favor). Insofar as the Court’s combined opinion resolves the appeal in People v Favor by not applying the Dokes (79 NY2d 656 [June 11, 1992]) superfluity exception and by giving the Dokes rule traditional retroactivity effect, I respectfully dissent and vote *269to affirm the order of the Appellate Division upholding defendant’s multiple convictions.
In the early morning hours of January 27, 1989, defendant seized a teenage girl, put a knife to her neck, ordered her to proceed down a driveway, and raped her. The victim described her attacker to the police, and after the perpetrator was arrested, she identified him in a lineup. Defendant was indicted for five crimes, including rapes against two different victims. Based on the identification of one of the victims and her testimony, defendant was convicted, after a jury trial on severed counts, of forcible rape in the first degree and kidnapping in the second degree. He did not testify. After his conviction, he pleaded guilty to the other rape, kidnapping and attempted robbery counts.
On the morning before the voir dire selection of the jury for trial, the trial court discussed the Sandoval issue in chambers with the prosecutor and defendant’s counsel. Since it was an untranscribed in-chambers conference, the record is silent as to whether defendant was present, but it appears uncontested that he was absent. However, with defendant present later that day in the courtroom, the trial court then recited for the record the result of the morning chambers’ conference:
"the court: All right. Before we started selecting the jury this morning we had a Sandoval conference in chambers. And I was told the defendant had a conviction for attempted unauthorized, or excuse me attempted unlawful imprisonment in the second degree from 1986 to which was sentenced to probation. I indicated that based on the similarity of unlawful imprisonment to kidnapping in the second degree that I would not allow the people to ask about that conviction’s underlying facts or the title of that conviction, but they could ask if he had ever been convicted of a misdemeanor. Are there any other issues you’d like to put on the record at this point?
"mr. Jacobs: Just that the district attorney informed me of no other bad acts which did not result in conviction. That was the only thing that I was aware of.
"mr. schauseil: Right.
"the court: Okay. Anything else before we have the openings tomorrow?” (Emphasis added.)
*270That is the entire record on the Sandoval issue. Defense counsel made no objection to the overwhelmingly favorable defense ruling, nor did counsel or defendant, at the express invitation of the trial court, request review or reopening of the bare misdemeanor aspect or any other aspect. The Appellate Division affirmed the conviction (172 AD2d 1052), and upon leave by a Judge of this Court, the Court reverses the convictions.
The majority retroactively applies the Dokes absolute-presence-no-preservation rule and holds that the Sandoval procedure employed by the Trial Judge in 1989, three years before Dokes was decided, nevertheless automatically entitles defendant to a new trial. The sole item with respect to which the trial court would have allowed cross-examination of defendant, had he chosen to testify, was the barest official fact and matter of record — his prior conviction for a misdemeanor. The ruling interdicted the People from identifying the nature, title or underlying facts of that prior misdemeanor. There is no cognizable or meaningful way defendant’s presence at the in-chambers discussion could have materially contributed to the trial court’s Sandoval ruling. To suggest otherwise on this record is speculative. Thus, the pro forma "absence” of defendant, as is customary (though no longer in Sandoval matters), from the Trial Judge’s in-chambers conference with the lawyers — for security and other logistical reasons — was not an "exclusion.” Also, it was truly "superfluous” on the sparse record, within the Dokes exception and within the comparable analysis applied to affirm the companion case here, People v Smith (decided today). I do not believe the labored distinctions between Favor and Smith on superfluousness hold up on close analysis.
In addition to my superfluousness analysis concerns, I believe the distinctions between Favor and People v Mitchell (80 NY2d 519 [Dec. 17, 1992]) are also unavailing. I discern no principled basis for rejecting in Favor the Mitchell (supra) prospectivity approach in its handling of the People v Antommarchi (80 NY2d 247 [Oct. 27, 1992]) presence rule. Prospectivity is before this Court for the first time in Favor, and is cogently presented for critical resolution based on recent precedential authority of this Court — Mitchell. By acknowledging the significantly expanded procedural and substantive features of Dokes and its sequelae and parallel procedural categories, this Court could quite justifiably utilize the analytical channel provided by People v Pepper (53 NY2d 213, cert *271denied. 454 US 967) to apply the Dokes formulation only prospectively (see, Cardozo, The Nature of the Judicial Process, reprinted in Selected Writings of Benjamin Nathan Cardozo, at 170 [Margaret E. Hall ed 1947]). It is no answer, I respectfully submit, that this problem, unlike the State-wide Antommarchi (80 NY2d 247, supra) impact, is claimed to be confined to Fourth Department practice and cases. The fact is otherwise, and the number of automatic reversals is very large and significant. The result unnecessarily unsettles hundreds of tried cases across the State. Thus, to me, Mitchell is to Antommarchi as Favor ought to be to Dokes.
I am mindful of the circumstance here that, unlike the coincidental back-to-back appeals and sessions of Court in Antommarchi and Mitchell, the Court is confronted in People v Favor with a relatively short period of time during which several appeals (see, e.g., People v Cruz, 81 NY2d 738; People v Gebrosky, 80 NY2d 995; People v Alexander 80 NY2d 801) were decided by this Court shortly after Dokes by applying its new rule. That rapid sequence feature is a distinction without legal significance and should not foreclose the Pepper prospectivity approach, especially since the Court correspondingly recognizes in this case that the retroactivity versus prospectivity issue was not argued in any of those intervening cases.
The Court has principled, analytical and precedential bases, in my view, to treat the intervening cases as serendipitous interludes and coincidences, which do not preclude application of the Antommarchi-Mitchell matrix to the parallel prospectivity problem presented here. The practical consequences in countless cases are equally devastating, and the rules in Dokes (79 NY2d 656, supra [defendant’s right to be present at a Sandoval hearing]) and Antommarchi (80 NY2d 247, supra [defendant’s right to be present at sidebar discussions]) certainly have sufficient related characteristics to warrant like prospective treatment.
Moreover, I am unpersuaded by that portion of Judge Titone’s opinion that concludes that this Court did not break "any new legal ground” in Dokes. Interestingly, he refers to mere "foreshadowings” in prior law for the characterization that Dokes is merely a "point in a continuum” (majority opn, at 264, 265), despite the fact that prior to Dokes this Court had not previously addressed the right to be present at Sandoval hearings, to say nothing of its newly promulgated forgiveness of the usual preservation requirements.
*272The proposition that Dokes broke new ground and broadened the presence-of-defendant procedural regimes is, in my view, unassailable. Simply put, the new rule insures automatic new trials in countless cases (1) without the necessity of any of this Court’s usually strict preservation requirements, a general protocol reasonably relied upon by the Bench and Bar (no objection is required and even tactical silence does not bar the ultimate relief); (2) without any examination or showing of prejudice or harmless error analysis (the superfluity exception mentioned in the new Dokes rule is in this case effectively rendered de minimis and different in kind, not just grade, from the genre of prejudice); and (3) with full retroactivity.
This Court’s expectation of clairvoyance from Trial Judges in anticipating how this Court eventually ruled on this issue, years before we did so, is not well founded. No Trial Judge could have reasonably foreseen these ratchetings. Yet, it is Trial Judges who ultimately bear the responsibility for these new trials that are being automatically mandated. They suffer these rebuffs to their handling of numerous trials because they were not prescient enough, despite otherwise fastidious precautions and procedures employed in the light of the "presence-preservation” law as it was before 1992. Protection of the record and the defendants’ rights, too, as in this 1989 case, with an almost entirely favorable defense ruling on the Sandoval issue itself, is held insufficient.
It cannot be overlooked that this case presents yet another category of cases in which virtual per se results are compelled, and now fully retroactively. The relatively recent phenomenon of this Court’s expanding per se jurisprudence appears to contradict the Holmesean-Cardozean notion of interstitial, incremental development in the common-law decisional process (Cardozo, The Nature of the Judicial Process, reprinted in Selected Writings of Benjamin Nathan Cardozo, at 134, 150 [Margaret E. Hall ed 1947]). Facets of individual cases intended to achieve the individually weighed justice desideratum are lost when automatic applications and results are commanded without regard to a broader-based examination of all relevant considerations and policies in each case.
In sum, my vote to affirm in Favor is driven by the Court’s unnecessarily restrictive construction of the Dokes superfluous presence exception for the first time on the record of the trial court’s ruling in this case. This construction and application is theoretically, functionally, practically and effectively indistin*273guishable from the analysis applied in People v Smith (decided today), yet with diametrically opposite results. Here, the Court uses the pure per se sweep. The starkness of this realization has led me to question the correctness of my own vote in Dokes. However, I cannot take that back and must acknowledge its stare decisis impact. That does not mean I cannot question the application of the deminimized superfluousness exception or look for a legitimate jurisprudential palliative, as the entire Court did in People v Mitchell (80 NY2d 519, supra), after recognizing and reciting the devastating consequences that per se nonpreservation presence rule would have if it were given retroactive operative effect. If this Court could make the sidebar presence rule applicable only to trials after the Antommarchi date, I cannot fathom why it cannot do the same thing here with this Dokes’ Sandoval presence rule, which inflicts comparably devastating consequences. The numerous automatic reversals recited by the majority itself, occurring in this Court alone, are testament to the widespread debacle and phenomenon in all the lower courts (see also, amicus curiae brief of NY State Dist Attorneys Assn).
My vote is to affirm the order of the Appellate Division upholding the conviction.
In People v Favor: Order reversed and a new trial ordered.
Chief Judge Kaye and Judges Simons, Hancock, Jr., Smith and Levine concur with Judge Titone; Judge Bellacosa dissents and votes to affirm in a separate opinion.
In People v Smith: Order affirmed.
Chief Judge Kaye and Judges Simons, Hancock, Jr., Bellacosa, Smith and Levine concur.