Summary
holding that an error is prejudicial "if an appellate court concludes that there is a significant probability, rather than only a rational possibility, in the particular case that the jury would have acquitted the defendant had it not been for the error or errors which occurred"
Summary of this case from Strzelecki v. CunninghamOpinion
Argued January 8, 1975
Decided February 25, 1975
Appeal from the Appellate Division of the Supreme Court in the Second Judicial Department, GEORGE J. BALBACH, J.
Nicholas Ferraro, District Attorney (Thomas A. Duffy, Jr. and Barry Alan Schwartz of counsel), for appellant.
William M. Erlbaum, Herbert A. Lyon and Charles Wender for respondent.
On this appeal we are called on principally to consider the doctrine of harmless error as applied to errors which occurred on defendant's second trial. In this case a mother was charged with criminal responsibility in connection with the deaths of her son and her daughter. On her first trial defendant was charged only with the death of the daughter and was convicted of manslaughter. On appeal this conviction was reversed and a new trial was ordered. ( People v. Crimmins, 33 A.D.2d 793, affd. 26 N.Y.2d 319. ) On her second trial the jury convicted defendant of murder of her son and manslaughter of her daughter. The Appellate Division then reversed the conviction of murder of the son and dismissed the charge against defendant with respect to his death. ( People v. Crimmins, 41 A.D.2d 933.) As to the manslaughter conviction, the Appellate Division also reversed defendant's conviction but ordered a new trial with respect to her responsibility for the death of her daughter. The case is now before us on appeal by the People.
The procedural aspects of this appeal and of our dispositions of its several branches call for exposition. The ultimate issues turn on the procedural significance and consequences properly to be attached to errors of law which occurred during the second trial. We conclude that these errors fall into separate categories calling for different legal results.
I. As to defendant's conviction of murder of her infant son:
The Appellate Division's reversal of this conviction (as distinguished from that court's attendant dismissal of this count in the indictment) was explicitly recited to be "on the law and the facts". An appeal may be taken to our court only where the reversal is expressly stated to be on the law alone; accordingly an appeal from this reversal may not be taken to our court (CPL 450.90, subd. 2, par. [a]).
By contrast, the corrective action directed by the Appellate Division in consequence of its reversal of the murder conviction, i.e., the dismissal of the murder count, is subject to an appeal to and review by our court (CPL 450.90, subd. 2, par. [b]). We find that corrective action to have been what was required by the Criminal Procedure Law. The reversal of the conviction was based on the conclusion of the Appellate Division that, as a matter of law, the People did not prove that the son's death resulted from a criminal act and, in the alternative, that any finding that it did would be contrary to the weight of the evidence ( 41 A.D.2d 933). CPL 470.20 (subd. 2) mandates dismissal of the accusatory instrument in the event of reversal of a judgment after trial for legal insufficiency of trial evidence; subdivision 5 of the same section mandates the same corrective action where the reversal is on the ground that the verdict is against the weight of the trial evidence. Accordingly the Appellate Division's dismissal of the murder count with respect to the death of the son must be affirmed.
II. As to defendant's conviction of manslaughter in the homicide of her infant daughter:
The Appellate Division determined that because of errors committed on the second trial this conviction should be reversed. Because such determination was expressly stated to be on the law alone, that aspect of the present appeal, as well as the associated corrective action directed by the Appellate Division, is properly before us (CPL 450.90, subd. 2, pars. [a], [b]). For reasons discussed below, a majority of our court is of the view that this determination of the Appellate Division should itself be reversed. In that circumstance, since the order of the Appellate Division reversing the manslaughter conviction was based on the law alone, the provisions of CPL 470.40 (subd. 2, par. [b]) dictate that the manslaughter conviction be remitted to the Appellate Division for determination of the facts. Presumably consideration will then be revived, too, as to defendant's separate and distinct appeal from the order of Supreme Court denying her motion for a new trial on the grounds of newly-discovered evidence and of improper conduct by the prosecutor in withholding from defendant information potentially helpful to her defense. In view of the other determinations made at the Appellate Division in the order from which appeal is now being taken it was not then necessary formally to reach or dispose of defendant's contentions with respect to denial of her motion for a new trial. Defendant now becomes entitled to consideration and disposition of such contentions by that court.
We turn then to a discussion of our reasons for concluding that the reversal of the manslaughter conviction should be overturned.
A. As to the constitutional error:
The People concede that the comment of the prosecutor in summation with respect to defendant's failure to testify on her own behalf was improper and constituted constitutional error under the provisions of both the Federal and our State Constitutions (U.S. Const., 5th Amdt.; N.Y. Const., art. I, § 6). All of the members of the court agree that such error calls for reversal and a new trial unless it was harmless under the test for harmless constitutional error laid down by the Supreme Court of the United States, namely, that there is no reasonable possibility that the error might have contributed to defendant's conviction and that it was thus harmless beyond a reasonable doubt ( Chapman v. California, 386 U.S. 18; Fahy v. Connecticut, 375 U.S. 85).
We of the majority are satisfied that this test is met here in view of the circumstances in which the constitutional error occurred — inter alia, the unsworn outbursts by defendant herself which both preceded and followed the prosecutor's error, the comments of defense counsel and the reactions in the courtroom at the time, and the explicitly clear instructions of the trial court — coupled with what, as indicated below, we think was the overwhelming proof of defendant's guilt.
Although in our view this case presents no appropriate instance for its application, our discussion of the effect to be given constitutional error should not overlook a parallel, and in some instances an overlapping doctrine, also of constitutional proportion, namely, the right to a fair trial. Not only the individual defendant but the public at large is entitled to assurance that there shall be full observance and enforcement of the cardinal right of a defendant to a fair trial. The appellate courts have an overriding responsibility, never to be eschewed or lightly to be laid aside, to give that assurance. So, if in any instance, an appellate court concludes that there has been such error of a trial court, such misconduct of a prosecutor, such inadequacy of defense counsel, or such other wrong as to have operated to deny any individual defendant his fundamental right to a fair trial, the reviewing court must reverse the conviction and grant a new trial, quite without regard to any evaluation as to whether the errors contributed to the defendant's conviction. The right to a fair trial is self-standing and proof of guilt, however overwhelming, can never be permitted to negate this right. There is no predicate here, however, for any claim that this defendant on her second trial was deprived of any such basic right.
B. As to the nonconstitutional errors:
For the purposes of our disposition of this appeal we assume, although each of the Judges in the majority would not necessarily so decide, that the Appellate Division was correct in concluding that in the circumstances of this trial: (1) it was error to permit introduction of testimony with respect to the witness Rorech's having been given a sodium pentothal (truth serum) test (although nothing was said as to any results thereof); (2) it was error to permit the prosecutor to elicit testimony in cross-examination of defendant's witness Colabella that the latter had refused to sign a waiver of immunity when questioned by the prosecutor during the pretrial investigation of the case; and (3) it was error, after the prosecutor had put before the jury an apparently damaging admission by Colabella to one Sullivan but had thereafter failed to call Sullivan or to explain the failure to do so, for the trial court to deny defendant's request for a charge that the jury could draw an unfavorable inference from the People's failure to call Sullivan as a witness. None of these errors, however, was of constitutional dimension.
We turn then to the question whether any one of such errors, or all taken in combination, calls for a reversal of the jury verdict here.
The definition and elaboration of the doctrine of harmless error as applied to nonconstitutional error involve peculiarly questions of the law of the State of New York to be determined by our State courts. The doctrine has received expression in our court over the last 20 years in various forms, accompanied usually explicitly, always at least implicitly, by a recognition that "[e]rrors are almost inevitable in any trial, improprieties almost unavoidable, [and that] the presence of one or the other furnishes no automatic signal for reversal and retrial" ( People v. Kingston, 8 N.Y.2d 384, 387).
Examination of the language chosen to describe the doctrine and its application in individual cases, as well as analysis of the authorities selected for citation, discloses that we have not always been either consistent in our classification or uniform in our expression. Forms of our verbalization of the doctrine cannot be nicely harmonized. Often there has been no expilcit recognition that there is a distinction between constitutional and nonconstitutional error; citations and verbiage have frequently been indiscriminately interchanged. On the other hand, we have never expressly held, as the dissent now urges, that there is no difference in the application of the doctrine of harmless error between constitutional and nonconstitutional error. When we have reached the conclusion that the error was harmless we have stated the rule loosely, in terms relatively easily satisfied. On the other hand when we have concluded that the error was not harmless our statement has been of a tight, demanding rule. The ultimate result in the individual case has been more significant than the particular formulation of the rule. (For cases decided in recent years see People v. Brosnan, 32 N.Y.2d 254, 262; People v. Stanard, 32 N.Y.2d 143, 148; People v. Steiner, 30 N.Y.2d 762, 763-764; People v. Crimmins, 26 N.Y.2d 319, 324-325; People v. Baker, 26 N.Y.2d 169, 174; People v. McKinney, 24 N.Y.2d 180, 185; People v. Pelow, 24 N.Y.2d 161, 167; People v. Miles, 23 N.Y.2d 527, 544; People v. Mirenda, 23 N.Y.2d 439, 446-447; People v. Cefaro, 23 N.Y.2d 283, 290; People v. Savino, 22 N.Y.2d 732, 733; People v. Adams, 21 N.Y.2d 397, 402; People v. Fein, 18 N.Y.2d 162, 175; People v. Donovan, 13 N.Y.2d 148, 153-154; People v. Duncan, 13 N.Y.2d 37, 42; People v. Rosenfeld, 11 N.Y.2d 290, 299-300; People v. Rosario, 9 N.Y.2d 286, 290-291; People v. Steinhardt, 9 N.Y.2d 267, 271-272; People v. Kingston, 8 N.Y.2d 384, 387; People v. Jackson, 7 N.Y.2d 142, 145; People v. Dziobecki, 3 N.Y.2d 997, 999; People v. Ochs, 3 N.Y.2d 54, 57; People v. Savvides, 1 N.Y.2d 554, 557, 558; People v. Mleczko, 298 N.Y. 153, 162-163.)
The presently applicable legislative statement of our State's rule, like its predecessor, has not been helpful. "An appellate court must determine an appeal without regard to technical errors or defects which do not affect the substantial rights of the parties" (CPL 470.05, subd. 1). The choice of the adjective "technical" in referring to errors may be said to connote those of a formalistic or minor character. On the other hand, to refer to errors which may affect "substantial" rights suggests errors of a somewhat more serious nature. Notably there has never been incorporated in the statutory language any concept of "harmlessness beyond a reasonable doubt". In any event, our decisions have not turned on or even been significantly affected by the legislative diction of present CPL 470.05 (subd. 1) or of section 542 of the former Code of Criminal Procedure.
It is appropriate therefore to recognize and to delienate the difference between the Federal harmless error rule with respect to constitutional error and our State's harmless error rule with respect to nonconstitutional error.
Two discrete considerations are relevant and have combined in varying proportions to produce specific results in particular cases. The first of such factors is the quantum and nature of proof of the defendant's guilt if the error in question were to be wholly excised. The second is the causal effect which it is judged that the particular error may nonetheless have had on the actual verdict. It appears that it is the latter consideration which is critical in the application of the Supreme Court test as to harmlessness of constitutional error. Thus, however overwhelming may be the quantum and nature of other proof, the error is not harmless under the Federal test if "there is a reasonable possibility that the * * * [error] might have contributed to the conviction" — perhaps the most demanding test yet formulated ( Fahy v. Connecticut, 375 U.S. 85, 86, supra; Chapman v. California, 386 U.S. 18, supra).
There may be other identifiable considerations in special instances (e.g., People v. Savvides, 1 N.Y.2d 554, supra) in which our court held the conduct of the prosecutor to be so improper as to call for a new trial quite irrespective either of the quantum of evidence of guilt or of any evaluation of the actual effect of the misconduct, at least in part for therapeutic purposes.
Our State rule to determine harmlessness of nonconstitutional error is not the same as the Federal rule.
The ultimate objective, grounded in sound policy considerations, is the wise balancing, in the context of the individual case, of the competing interests of the defendant and those of the People. "While we are ever intent on safeguarding the rights of a defendant * * * we recognize at the same time that the State has its rights too" ( People v. Kingston, 8 N.Y.2d 384, 387, supra). Thus, it does not follow that an otherwise guilty defendant is entitled to a reversal whenever error has crept into his trial. On the other hand, we recognize that a finding that an error has not been harmless does not result in fatal consequences to the People; they are put to a new trial, but the defendant does not go free.
Our State test with respect to nonconstitutional error is not so exacting as the Supreme Court test for constitutional error. We observe that in either instance, of course, unless the proof of the defendant's guilt, without reference to the error, is overwhelming, there is no occasion for consideration of any doctrine of harmless error. That is, every error of law (save, perhaps, one of sheerest technicality) is, ipso facto, deemed to be prejudicial and to require a reversal, unless that error can be found to have been rendered harmless by the weight and the nature of the other proof. That "overwhelming proof of guilt" cannot be defined with mathematical precision does not, of course, mean that the concept cannot be understood and applied in individual cases, although not always without some difficulty. It surely does not invite merely a numerical comparison of witnesses or of pages of testimony; the nature and the inherent probative worth of the evidence must be appraised. As with the standard, "beyond a reasonable doubt", recourse must ultimately be to a level of convincement. What is meant here, of course, is that the quantum and nature of proof, excising the error, are so logically compelling and therefore forceful in the particular case as to lead the appellate court to the conclusion that "a jury composed of honest, well-intentioned, and reasonable men and women" on consideration of such evidence would almost certainly have convicted the defendant.
If, however, an appellate court has satisfied itself that there was overwhelming proof of the defendant's guilt, its inquiry does not end there. Under our system of justice a jury is not commanded to return a verdict of guilty even in the face of apparently conclusive proof of the defendant's guilt. Similarly it may and often does exercise a positive sense of moderating mercy. Further inquiry must accordingly be made by the appellate court as to whether, notwithstanding the overwhelming proof of the defendant's guilt, the error infected or tainted the verdict. An evaluation must therefore be made as to the potential of the particular error for prejudice to the defendant. We hold that an error is prejudicial in this context if the appellate court concludes that there is a significant probability, rather than only a rational possibility, in the particular case that the jury would have acquitted the defendant had it not been for the error or errors which occurred.
Turning then to the record now before us, we of the majority conclude that, excising both the evidence erroneously admitted (with respect to Rorech's taking a truth test and as to Colabella's refusal to sign a waiver of immunity) and the prosecutor's interrogation of Colabella (as to the latter's damaging admission to Sullivan), there was overwhelming proof that this defendant was guilty of manslaughter in the death of her daughter. In addition to other compelling circumstancial evidence, there was eyewitness testimony (unavailable to support the conviction on the first trial because it had been infected by the wholly improper visit of jurors to the scene) that on the night before the daughter's body was found, defendant, carrying what was described as a "bundle" and accompanied by an unidentified man, was seen leading her son from the Crimmins home; that as the man threw the "bundle" into a parked car defendant cried out, "Please don't do this to her", to which the man responded, "Does she know the difference now? * * * Now you're sorry." Additionally defendant herself later confessed her guilt to her paramour — "Joseph, forgive me, I killed her." On the other hand the description which defendant offered of the events of the evening preceding the children's disappearance was completely discredited and the prosecution conclusively exploded defendant's theory of an outside kidnapper. We read this record as leading only to a single, inexorable conclusion, as two juries have indeed found: defendant was criminally responsible for the death of her daughter.
Proceeding further, then, as we must, we also conclude that in the circumstances of this case there is no significant probability in the light of the overwhelming proof that, had it not been for the errors which occurred, this jury would have acquitted the defendant or that a third jury might do so. Our ultimate conclusion, therefore, is that under our State rule the nonconstitutional errors which occurred on this defendant's second trial were harmless.
The order of the Appellate Division with respect to the manslaughter conviction should accordingly be reversed, and the case remitted to the Appellate Division for determination of the facts in conformity with CPL 470.40 (subd. 2, par. [b]).
I
I agree with the court's disposition of the appeal from that portion of the order of the Appellate Division which reversed defendant's conviction of the murder of her infant son and dismissed that count of the indictment.
II
With respect to the manslaughter count, I would affirm the order of the Appellate Division. There is reason for grave concern because of the rule formulated by the majority for the review of "nonconstitutional" errors and its application to this case.
As to errors of constitutional dimension, the majority recognizes the standard of Chapman v. California ( 386 U.S. 18, 24), that before a constitutional error can be held harmless the court must be able to declare a belief that it was harmless beyond a reasonable doubt. This standard is followed by the declaration of "a parallel, and in some instances an over lapping doctrine, also of constitutional proportions, namely, the right to a fair trial," such that "if in any instance, an appellate court concludes that there has been such error of a trial court, such misconduct of a prosecutor, such inadequacy of defense counsel, or such other wrong as to have operated to deny any individual defendant his fundamental right to a fair trial, the reviewing court must reverse the conviction and grant a new trial, quite without regard to any evaluation as to whether the errors contributed to the defendant's conviction" (emphasis supplied) (p. 238). As to "nonconstitutional errors", the majority establishes (p. 242) the further precept that, if "an appellate court has satisfied itself that there was overwhelming proof of defendant's guilt," further inquiry must be made by it "as to whether * * * the error infected or tainted the verdict" and "an error is prejudicial in this context if the appellate court concludes that there is a significant probability, rather than only a rational possibility, in the particular case that the jury would have acquitted the defendant had it not been for the error or errors which occurred."
While the effort to harmonize the Constitutions, the statute and judicial pronouncements and to render a yardstick to guide courts in the conduct of criminal trials and in reviewing alleged errors therein is creditable, it is urged respectfully that the majority's opinion does not accomplish that result. To begin with, what has evolved is indeed a trifurcated standard for appellate scrutiny. There is a fork of error "harmless beyond a reasonable doubt" as to "constitutional" deprivations, another "also of constitutional proportion, namely, the right to a fair trial * * * quite without regard to any evaluation as to whether the errors contributed to the defendant's conviction" (pp. 237-238) and a third with a test of "significant probability" applicable to nonconstitutional errors. This three-pronged measure will be difficult to administer and apply and, instead of clarity, confusion comes forth.
Although the decision of the Supreme Court in Chapman did not purport to establish a harmless error rule for application to all errors, there are strong reasons for applying the "harmless beyond a reasonable doubt" standard to all errors affecting the substantial rights of a party which arise under the State Constitution or State law, as well as to those which emanate from the Federal Constitution. Such a rule would not be inconsistent with the mandate of CPL 470.05 (subd. 1), which merely directs an appellate court to determine an appeal without regard to technical errors which do not affect the substantial rights of the parties.
First, if the nature of the error is constitutional, it is going to be difficult, if not impossible, to determine which test shall be applied. Should it be according to the Chapman scale where "the court must be able to declare a belief that it [the constitutional error] was harmless beyond a reasonable doubt"? Or, should there be a weighing to ascertain if there has been a "full observance and enforcement of the cardinal right of a defendant to a fair trial * * * quite without regard to any evaluation as to whether the errors contributed to the defendant's conviction"? (p. 238). Although it is obvious that the criteria are not the same, the majority does not supply the answer.
Second, to establish a coexisting rule that where there has been denied to "any individual defendant his fundamental right to a fair trial, the reviewing court must reverse the conviction and grant a new trial, quite without regard to any evaluation as to whether the errors contributed to the defendant's conviction" (p. 238) is incongruous, since generally one of the most significant inquiries to be made in ascertaining whether a fair trial has been accorded a defendant concerns the effect of any error, misconduct, inadequacy or wrong upon the verdict. Such a rule would unnecessarily place in jeopardy a host of convictions.
Third, it is apparent from a reading of Chapman v. California ( 386 U.S. 18, 24, supra) that the test of "harmless beyond a reasonable doubt" was adopted as a corollary to the reasonable doubt standard applicable to criminal cases. By adopting a test regarding so-called "nonconstitutional" errors which requires "a significant probability * * * that the jury would have acquitted defendant had it not been for the error or errors which occurred" (p. 242), the court is dangerously diluting the time-honored standard of proof beyond a reasonable doubt which has been a cornerstone of Anglo-Saxon criminal jurisprudence. No one would dispute the statement that a defendant in a criminal case has a constitutional right to be proven guilty beyond a reasonable doubt before he is deprived of his life, liberty or property (U.S. Const., 5th Amdt., 14th Amdt., § 1; N.Y. Const., art. I; In re Winship. 397 U.S. 358, 363-364; La Fave Scott, Criminal Law, Hornbook Series, pp. 45-46; cf. Matter of Richard S., 27 N.Y.2d 802; see CPL 70.20). Unless an appellate court can say that errors committed at trial, which affected defendant's substantial rights, are harmless beyond a reasonable doubt, defendant's right to that standard of proof can be severely prejudiced, the extent of the prejudice depending upon the nature of the error in the context of other proof and the circumstances of the case.
Fourth, while the conceptual distinction between constitutional and nonconstitutional errors is a real one, the differentiation is of dubious validity as applied to the appellate review process. As the history of the "right-privilege" distinction in law indicates, the process of ascribing labels to concepts from which serious consequences flow is one fraught with peril and one uniquely susceptible to semantic gamesmanship. (For a history of the right-privilege distinction see Van Alstyne, The Demise of the Right-Privilege Distinction in Constitutional Law, 81 Harv. L. Rev. 1439.) Ultimately, the Supreme Court opted for a course of action where the consequences to the particular party, rather than the label attached, determined the scope of due process safeguards (see Goldberg v. Kelly, 397 U.S. 254, 262-263).
The pitfalls of utilizing the "constitutional" and "nonconstitutional" dichotomy to determine the standard of review are apparent. According to the rule enunciated by the majority, the "harmless beyond a reasonable doubt" test is applied where a constitutional error, other than one which denied the right to a fair trial, is involved; however, in a nonconstitutional error situation, the test of "significant probability * * * that the jury would have acquitted" (p. 242) but for the error comes into play. Given a case, like this one, where there is an accumulation of errors, or even in other situations involving a single error, the effect of which is such as to deprive a defendant of his or her constitutional right to a fair trial ( Irvin v. Dowd, 366 U.S. 717, 722; In re Murchison, 349 U.S. 133, 136), even the most intense student of the law must wonder which standard is to be applied. Of course, if a single standard of harmless error beyond a reasonable doubt is in force, such difficulties would not arise.
More fundamentally, a defendant's constitutional right to a fair trial can be prejudiced equally by nonconstitutional errors as by constitutional errors. This is why it makes little sense to have a strict standard of review in one area and a looser one in the other. In the instant matter, for example, the constitutional error, which triggers the more exacting standard of review, was not of great significance in the context of the trial and was provoked to some extent by defendant's action, as the majority recognizes. Far more serious, in my opinion, were the errors the majority labels as nonconstitutional. On this record, these and other errors cannot be shown to be harmless beyond a reasonable doubt. In any event, their effect was to deprive defendant of a fair trial and the conviction must be reversed (see People v. Trybus, 219 N.Y. 18, 21).
One of the major items of evidence received at the trial was an admission made by defendant to Joseph Rorech in which she admitted killing her daughter. The majority does not deny that it was error to permit introduction of testimony that Rorech, an important prosecution witness, had been given a sodium pentothal (truth serum) test. It is difficult to imagine evidence that could have had as grave an impact upon the jury in their assessment of the credibility of the witness Rorech, to defendant's obvious detriment. Under any view of the case, the error affected defendant's "substantial rights" (CPL 470.05, subd. 1) to have the jury evaluate the believability of the witness without the distraction of totally irrelevant considerations injected by evidence erroneously received.
The prosecution's theory was that Colabella was with defendant on the night her daughter was killed and helped her dispose of the body. The admission into evidence of proof regarding his failure to sign a waiver of immunity, when questioned by the prosecutor during the pretrial investigation of the case, was "most improper" and also affected defendant's "substantial rights" since, as observed by the majority in the Appellate Division, his refusal may well have been considered by the jury as an indication of defendant's guilt (cf. People v. Ashby, 8 N.Y.2d 238, 242-243; United States v. Sing Kee, 250 F.2d 236, 240, cert. den. 355 U.S. 954). The existence of this issue, involving indirectly at least the Fifth Amendment, has "constitutional overtones" (see Namet v. United States, 373 U.S. 179, 186-187; Grunewald v. United States, 353 U.S. 391, 423-424) and points up again the difficulty in assigning the labels employed by the majority.
Likewise, the seriousness of the error, in putting before the jury Colabella's alleged admission to Sullivan that he had a girl friend, Alice, "who was in a jam," and asking Sullivan to help get rid of a body, is readily apparent. Colabella denied not only knowing Sullivan but making the admission, and Sullivan was not called as a witness nor was an explanation given for the failure to call him. As pointed out in the decision under review, this impropriety was aggravated when "(a) the prosecutor strenuously opposed a subsequent defense request for a charge that the jury could draw an unfavorable inference from the People's failure to call Sullivan as a witness and (b) the court refused to so charge" ( 41 A.D.2d 933).
The trial was infected with further prejudicial error in placing before the jury the subject of defendant's trip to the Bahamas with a married man. The only conceivable relevancy of this item was on the question of motive but testimony bearing on this subject would be incompetent since there was not a logical relation between it and the commission of the crime charged "according to known rules and principles of human conduct" ( People v. Fitzgerald, 156 N.Y. 253, 258; Richardson, Evidence [Prince — 10th ed.], § 171). As noted recently in People v. Sandoval ( 34 N.Y.2d 371, 376), "it must be recognized as inevitable * * * that evidence of prior criminal, vicious or immoral conduct will always be detrimental to the defendant."
Further difficulty is encountered regarding the introduction of evidence that, on the date of her alleged confession, defendant became extremely distraught upon reading a newspaper account of the arrests of 13 persons at the "Little Appalachian" meeting in Queens and kept repeating a name. It turned out later that Colabella was not named in the news story nor was it his name which defendant repeated. This Mafia "angle", introduced without relevance, was prejudicial and may have played an important part in the trial. One major witness, Sophie Earomirski, who allegedly saw defendant from a distance on the night of the crime carrying a bundle and overheard an incriminating conversation, justified her failure to come forward promptly with her evidence on the grounds of fear. With the prosecutor's injection of a spurious organized crime aspect to the case, in itself dangerous, her explanation very likely gained a credence it might otherwise have lacked.
The majority (p. 242) resolves the problem by "excising both the evidence erroneously admitted (with respect to Rorech's taking a truth test and as to Colabella's refusal to sign a waiver of immunity) and the prosecutor's interrogation of Colabella (as to the latter's damaging admission to Sullivan)" and by finding overwhelming evidence of guilt. Performing such radical surgery on the evidence fails to recognize sufficiently the danger of improperly influencing or "tainting" the verdict by "harmless errors". Jurors, hearing the events unfolding in an emotion-charged atmosphere may very well impute greater importance to evidence erroneously received than is apparent by speculation of appellate courts, removed from the environment and reading cold print. Their assessment of such evidence may color their entire outlook of defendant's legal position.
More importantly, however, it is not for this court to usurp the function of the jury and speculate whether, without this evidence erroneously admitted, the jury nevertheless would have acquitted (see dissent in People v. Catalanotte, 36 N.Y.2d 192). What this court wrote in People v. Marendi ( 213 N.Y. 600, 619) many years ago is just as true today: "where prejudicial matter is erroneously received in evidence on a disputed question of fact, its harmful character cannot be determined solely by the mere weight of competent evidence unless we are to resolve ourselves into a jury and, ignoring the finding upon incompetent evidence, substitute one upon the evidence which we may deem competent." That two juries have found guilt is beside the point, just as is the fact that the both verdicts, up to this point, have been set aside by different courts. This court, following the first trial said ( 26 N.Y.2d 319, 324): "Although, as the People argue, the evidence is legally sufficient to sustain the verdict of guilt, it was not so overwhelming that we can say, as a matter of law, that the error [then under review] could not have influenced the verdict ( Harrington v. California, 395 U.S. 250; Chapman v. California, 386 U.S. 18)." Emotions aside, the failure to accord an accused a fair hearing violates even the minimal standards of due process regardless of the heinousness of the crime charged (see Irvin v. Dowd, 366 U.S. 717, 722, supra).
The fact is that different items of prejudicial matter were admitted and were before the jury for its consideration. The character of the evidence was such that they may well have affected the jury's evaluation of other items of evidence, to defendant's detriment. Where these several major elements of evidence against defendant were tainted by error, it cannot be said that the case against defendant, although persuasive, was overwhelming and that a jury composed of honest, well-intentioned, and reasonable men and women could not have acquitted her.
I would affirm the order of the Appellate Division.
Chief Judge BREITEL and Judges JASEN, GABRIELLI and WACHTLER concur with Judge JONES; Judge COOKE concurs in part and dissents in part and votes to affirm in a separate opinion in which Judge FUCHSBERG concurs.
Order modified and case remitted to Appellate Division, Second Department, for further proceedings in accordance with the opinion herein and, as so modified, affirmed.