People v. Banks , 438 Mich. 408 ( 1991 )

438 Mich. 408 (1991) 475 N.W.2d 769

PEOPLE
v.
BANKS

Docket No. 86945, (Calendar No. 12).

Supreme Court of Michigan.

Argued April 4, 1991. Decided September 9, 1991.

Frank J. Kelley, Attorney General, Gay Secor Hardy, Solicitor General, John D. O'Hair, Prosecuting Attorney, Timothy A. Baughman, Chief, Research, Training and Appeals, and Joseph A. Puleo, Assistant Prosecuting Attorney, for the people.

Carl Ziemba for the defendant.

GRIFFIN, J.

The question presented in this case is whether the trial court erred in permitting the prosecutor to introduce into evidence at a joint trial, with limiting instructions, the redacted statements of two nontestifying codefendants and, if so, whether the error was harmless with regard to this defendant. We hold that admission of the redacted statements denied defendant his right of confrontation guaranteed by US Const, Am VI and Const 1963, art 1, § 20. The error was not harmless beyond a reasonable doubt when considered in the context of other properly admitted evidence. We therefore reverse the decision of the Court of Appeals and remand this case to the trial court for a new trial.

I

Defendant Melvin Banks was convicted of first-degree murder, MCL 750.316; MSA 28.548, possession of a firearm during the commission of a felony, MCL 750.227b; MSA 28.424(2), and three counts of assault with intent to commit murder, MCL 750.83; MSA 28.278, in the shooting death of Leonard Ingram, a Redford high school student, and the armed attack on three of his companions, one of whom was wounded.

Defendant Banks was tried jointly with two codefendants, Theodore Burley and Aaron Funches.^[1] The prosecution's theory was that Banks was the person who fired the gun, aided and abetted by Funches and Burley. The prosecutor submitted that the gun in question belonged to Funches, that Burley handed it to the defendant and pointed out Ingram and his friends, and that the defendant then confronted and shot at them.

Funches and Burley chose not to testify at trial. However, each of them had given the police a statement. A redacted version of their statements was read to the jury.^[2] In each instance, the word "blank" was substituted for the name of Banks and for the name of the other codefendant.^[3] The original or unredacted written statements of Funches and Burley were not introduced into evidence. The trial court instructed the jurors that each of the statements as read in open court was to be considered only in deciding the culpability of the person who gave the statement, and was not to be used in determining the guilt or innocence of any other defendant.

The defendant, Banks, testified in his own behalf. He asserted that although he had been in Funches' car, and later at Funches' home on the day in question, he had not been at the shooting site and was not involved in the shooting. He contended that he had met Burley for the first time that day at Funches' house. Defendant's counsel also argued misidentification, emphasizing that there had been no corporeal lineup, that none of the eyewitnesses had known the defendant previously, and that the witness who had participated in the photographic showup likely had seen the defendant's photograph in the newspaper.

The prosecution's witnesses included Ingram's three companions — Lawrence Jordan, Sean Davis, and Larry Harris. Each identified the defendant as the triggerman. None of the companions had known the defendant before the shooting, and Jordan and Davis admitted that they had learned his name from a newspaper article. Although Jordan picked out the defendant's photograph from an array of six during a showup, he admitted that he had previously seen the defendant's photograph in the newspaper. None of the victim's companions participated in a corporeal or voice lineup.

The jury acquitted codefendants Funches and Burley of all counts,^[4] but convicted the defendant. He was subsequently sentenced to serve a non-parolable term of life imprisonment for first-degree murder, parolable life terms for the three assault counts, and a mandatory two-year term for felony-firearm.

The Court of Appeals held that the nontestifying codefendants' statements had been admitted in violation of Bruton v United States, 391 U.S. 123; 88 S. Ct. 1620; 20 L. Ed. 2d 476 (1968). However, the panel was persuaded that the error was harmless. Unpublished opinion per curiam of the Court of Appeals, decided August 2, 1989 (Docket No. 106741).

Defendant's application for leave to appeal was granted by this Court, limited to the issues (1) whether the trial court erred in permitting the prosecutor to introduce into evidence at this joint trial the statements of two nontestifying codefendants, and (2) if so, whether the error was harmless.^[5] 435 Mich. 867 (1990).

II

The Sixth Amendment of the United States Constitution provides, in part, that "[i]n all criminal prosecutions, the accused shall enjoy the right ... to be confronted with the witnesses against him...." This Sixth Amendment right is applicable to the states, Douglas v Alabama, 380 U.S. 415; 85 S. Ct. 1074; 13 L. Ed. 2d 934 (1965), and the same right is guaranteed by Const 1963, art 1, § 20.^[6]

In California v Green, 399 U.S. 149, 158; 90 S. Ct. 1930; 26 L. Ed. 2d 489 (1970), the United States Supreme Court explained that the Confrontation Clause

(1) insures that the witness will give his statements under oath — thus impressing him with the seriousness of the matter and guarding against the lie by the possibility of a penalty for perjury; (2) forces the witness to submit to cross-examination, the "greatest legal engine ever invented for the discovery of truth"; (3) permits the jury that is to decide the defendant's fate to observe the demeanor of the witness in making his statement, thus aiding the jury in assessing his credibility.

In Pointer v Texas, 380 U.S. 400, 405; 85 S. Ct. 1065; 13 L. Ed. 2d 923 (1965), the Court observed:

There are few subjects, perhaps, upon which this Court and other courts have been more nearly unanimous than in their expressions of belief that the right of confrontation and cross-examination is an essential and fundamental requirement for the kind of fair trial which is this country's constitutional goal.

In Bruton v United States, supra, the Court held that a defendant is deprived of his Sixth Amendment right of confrontation when the facially incriminating unredacted confession of a nontestifying codefendant is introduced at their joint trial, even if the jury is instructed to consider the confession only against the codefendant. In Bruton, defendants Bruton and Evans were tried jointly for armed postal robbery. A postal inspector, the government's witness, testified regarding an oral confession allegedly made by Evans, which inculpated both Evans and Bruton. Evans did not testify. The trial court instructed the jury to disregard the confession in judging Bruton's guilt or innocence and to consider it only for the purpose of deciding Evans' culpability. The jury found both defendants guilty. The United States Court of Appeals for the Eighth Circuit affirmed, but the United States Supreme Court reversed, stating:

[T]here are some contexts in which the risk that the jury will not, or cannot, follow instructions is so great, and the consequences of failure so vital to the defendant, that the practical and human limitations of the jury system cannot be ignored.... Such a context is presented here, where the powerfully incriminating extrajudicial statements of a codefendant, who stands accused side-by-side with the defendant, are deliberately spread before the jury in a joint trial. Not only are the incriminations devastating to the defendant but their credibility is inevitably suspect, a fact recognized when accomplices do take the stand and the jury is instructed to weigh their testimony carefully given the recognized motivation to shift blame onto others. The unreliability of such evidence is intolerably compounded when the alleged accomplice, as here, does not testify and cannot be tested by cross-examination. It was against such threats to a fair trial that the Confrontation Clause was directed. [Bruton, supra, pp 135-136.]

In a footnote, the Bruton Court noted that other courts had tried to accommodate the interests of both the defendant and the prosecution by utilizing the process of redaction. The Court further noted, however, that redaction had been criticized as ineffective by some legal authorities. Id., p 134, n 10.

The Court subsequently had the opportunity to address the Bruton problem in the context of a redacted statement. In Richardson v Marsh, 481 U.S. 200; 107 S. Ct. 1702; 95 L. Ed. 2d 176 (1987), the Court held that the admission of a nontestifying codefendant's confession with a limiting instruction did not violate the Confrontation Clause when the confession was redacted to eliminate all references to the nonconfessing defendant.^[7] Defendant Marsh was convicted of felony murder and assault with intent to murder. The prosecution's theory was that defendant Marsh rode in the back seat of an automobile to the crime scene with her codefendant and a third person and, because she overheard the front seat occupants' conversation — a scheme to rob and kill the victims — she acquired the requisite intent. The prosecution offered the confession of the codefendant which stated that, during the ride, the codefendant and the driver formulated plans for the crime. Defendant Marsh admitted her presence in the back seat of the automobile, but denied that she heard the conversation because the car radio drowned out the front seat conversation. She testified that she was therefore surprised when her friends stole money from the victims and then shot them. The key issue in the case, the defendant's intent to commit the crime, thus turned on whether she had heard the conversation.

The prosecutor redacted the codefendant's confession to delete all references to Marsh not only by name but her existence as well. The trial court admitted the confession into evidence and gave the jury limiting instructions. The defendant argued that admission of the confession violated Bruton, supra, because the jury could infer from the confession that the defendant had heard the front seat conversation. However, the Supreme Court held that admitting the confession did not deny the defendant her right to confrontation. The Court distinguished Bruton, supra:

There is an important distinction between this case and Bruton, which causes it to fall outside the narrow exception we have created. In Bruton, the codefendant's confession "expressly implicat[ed]" the defendant as his accomplice. Id., at 124, n 1. Thus, at the time that confession was introduced there was not the slightest doubt that it would prove "powerfully incriminating." Id., at 135. By contrast, in this case the confession was not incriminating on its face, and became so only when linked with evidence introduced later at trial (the defendant's own testimony).

Where the necessity of such linkage is involved, it is a less valid generalization that the jury will not likely obey the instruction to disregard the evidence. Specific testimony that "the defendant helped me commit the crime" is more vivid than inferential incrimination, and hence more difficult to thrust out of mind.

* * *

The rule that juries are presumed to follow their instructions is a pragmatic one, rooted less in the absolute certitude that the presumption is true than in the belief that it represents a reasonable practical accommodation of the interests of the state and the defendant in the criminal justice process. On the precise facts of Bruton, involving a facially incriminating confession, we found that accommodation inadequate. As our discussion above shows, the calculus changes when confessions that do not name the defendant are at issue.... We hold that the Confrontation Clause is not violated by the admission of a nontestifying codefendant's confession with a proper limiting instruction when, as here, the confession is redacted to eliminate not only the defendant's name, but any reference to his or her existence. [Richardson, supra, pp 208-211.]

The Court rejected outright the alternative of separate trials whenever a codefendant's statement is sought to be used against him where it might also incriminate the defendant:

It would impair both the efficiency and the fairness of the criminal justice system to require, in all these cases of joint crimes where incriminating statements exist, that prosecutors bring separate proceedings, presenting the same evidence again and again, requiring victims and witnesses to repeat the inconvenience (and sometimes trauma) of testifying, and randomly favoring the last-tried defendants who have the advantage of knowing the prosecution's case beforehand. [Id., p 210.]

Ultimately, because the prosecutor in his closing argument may have undone the effect of the limiting instructions by urging the jury to consider the codefendant's confession in assessing Marsh's guilt, the Richardson Court remanded the case for consideration of whether the error could serve as the basis for granting a writ of habeas corpus, despite the defendant's failure to object to the prosecutor's comments. Id., p 211.

The Richardson Court expressly limited its ruling to those situations in which redaction eliminates not only the defendant's name, but also any reference to the defendant's existence or role in the crime:

We express no opinion on the admissibility of a confession in which the defendant's name has been replaced with a symbol or neutral pronoun. [Id., p 211, n 5.]

Since the Richardson holding stops short of the situation presented in this case, we are left to determine the effect of Bruton, when read in conjunction with Richardson, on a "partially" redacted statement.

The Richardson "facially incriminating" analysis does not easily lend itself to the present circumstances. Indeed, the line between inferential incrimination and direct implication is a thin one when the fact of a defendant's existence is not totally eliminated from a codefendant's statement. In such situations, where the defendant's name is merely replaced by a neutral pronoun or "blank," the statement may be "powerfully incriminating" within the meaning of Bruton, but not necessarily "facially incriminating" according to Richardson, because the jury must technically infer who "blank" might be. Certainly a jury can draw strong inferences even from a partially redacted confession if the confession is connected with other evidence at trial. The concerns voiced by Justice Stevens in his dissenting opinion in Richardson, supra, pp 213-214, are particularly germane to the present facts:

Today the Court nevertheless draws a line between codefendant confessions that expressly name the defendant and those that do not. The Court relies on the presumption that in the latter category "it is a less valid generalization that the jury will not likely obey the instruction to disregard the evidence." Ante, at 208. I agree; but I do not read Bruton to require the exclusion of all codefendant confessions that do not mention the defendant. Some such confessions may not have any significant impact on the defendant's case. But others will. If we presume, as we must, that jurors give their full and vigorous attention to every witness and each item of evidence, the very acts of listening and seeing will sometimes lead them down "the path of inference." Indeed, the Court tacitly acknowledges this point; while the Court speculates that the judge's instruction may dissuade the jury from making inferences at all, it also concedes the probability of their occurrence, arguing that there is no overwhelming probability that jurors will be unable to "disregard an incriminating inference." Ibid. Bruton has always required trial judges to answer the question whether a particular confession is or is not "powerfully incriminating" on a case-by-case basis; they should follow the same analysis whether or not the defendant is actually named by his or her codefendant. [Emphasis in the original.]

In cases such as the one before us now, where redaction is achieved through replacement of the defendant's name with a neutral pronoun or "blank," the ease with which a jury will be able to fill in a blank will vary from case to case, depending upon the overall evidentiary context in which it is introduced to the jury. A rule of admissibility per se is simply not appropriate for the form of redaction used in the case at bar.

We therefore return to the basic tenets of Bruton. Because codefendant statements are "inevitably suspect" because of the strong potential for blame shifting, Bruton, supra, p 136, even redacted confessions like the one challenged here should be clothed with a presumption of unreliability. If a "substantial risk" exists that the jury, despite cautionary instructions, will consider a codefendant's out-of-court statement in deciding the defendant's guilt, the statement — even though redacted to delete the defendant's name — will be rendered inadmissible at a joint trial. Bruton, supra, p 126. Other independent evidence may, by necessity, have to be considered:

[T]he court must decide whether the statement incriminates the defendant against whom it is inadmissible in such a way as to create a "substantial risk" that the jury will look to the statement in deciding on that defendant's guilt. Such an assessment may require consideration of other evidence in order to determine whether mere deletion of the defendant's name will be effective in making the statement non-incriminating as to him. But consideration of the weight of independent evidence is both improper and unnecessary to determination of the Bruton issue at the trial court level.[^[8]] [Hodges v Rose, 570 F2d 643, 647 (CA 6, 1978), cert den 436 U.S. 909 (1978). See also Foster v United States, 548 A2d 1370 (DC, 1988) (and cases cited therein).]

We believe that this case-by-case approach best protects the Confrontation Clause rights of a defendant in situations where redaction is used to edit an otherwise inadmissible out-of-court statement.

In the instant case, the prosecution charged that Funches and Burley aided and abetted the defendant, who allegedly did the shooting. Funches and Burley, through their statements, which were read to the jury, admitted that they were at the scene of the crime, but claimed that they had nothing to do with the shooting, and that they had no prior knowledge that defendant was going to shoot anyone. Burley's statement read in pertinent part:

Then the three of us me, blank and blank, drove off. We saw four boys walking on St. Marys going away from Grand River with their backs to us. Me said, yeah, it looks like them. As blank was turning the car, blank got the gun. Blank parked the car and he and I waited. Blank got out and walked up to the corner. He had his hand in his jacket pocket. He stood there for about a minute or so, said something, pulled out a gun and started shooting.

He fired about five shots, then ran back to the car and jumped in. Blank drove off. After we drove off I saw a towel in the glove box and I gave it to blank. It was in the front seat and blank was behind me in the rear seat.

I gave him the towel and he wiped the gun off.

Aaron Funches' statement read in pertinent part:

I heard one shot and then I saw blank running back to the car. He got in.

I didn't see it [the gun] on him. I saw the gun when he got back in the car. It was a small black gun like a .22.

One other thing I didn't mention was that when blank got out of the car to go around to where the guys were at on St. Mary's —

These statements directly contradicted defendant's testimony that he was elsewhere at the time of the shooting. Defendant claimed that he had gone to the high school on the afternoon in question to return some books. He had been enrolled in night school, but had decided to quit. Defendant left the school after about ten minutes, having learned from his younger brother that the teachers he wanted to see were not there. Defendant asked Funches for a ride home. Funches agreed, but said that he first had to pick up his sister. Funches drove across the street from the high school to wait for his girl friend. When she arrived, the group left to pick up Funches' sister, then went to Funches' home. The defendant testified that he met Burley for the first time while at Funches' house, and that he left Funches' residence at about 5:15 P.M., before the shooting occurred. He denied being at the site of the shooting or being involved in the shooting in any way.

We conclude that the process of redaction used in the instant case was ineffective in avoiding a Bruton error. Unlike Richardson, supra, the redaction here did not eliminate all reference to the existence of the defendant. To the contrary, the jury knew from both Funches' and Burley's statements that a third person had taken possession of a gun, had left the car while Burley and Funches remained in it, and had returned to the car after Burley and Funches heard shots. This "third person" and his actions were so described as to leave no doubt that it was the third person on trial — the defendant. The defendant rightfully complains that he might as well have been mentioned by name in each of the two statements by Burley and Funches.

Any lingering doubts regarding the transparency of the redaction are eliminated after a review of the closing arguments of the prosecutor and codefendants' counsel. As in Richardson, supra, p 211, the prosecutor in the instant case urged the jury to use the nontestifying codefendants' statements to evaluate the defendant's case, thereby effectively undoing the trial court's cautionary instructions. In his summation to the jury, the prosecutor argued:

Okay. Jerry Pringle is being dropped off and now left in the car is Melvin Banks, Aaron Funches, and Theodore Burley. Okay. What do we know what happens next? The statement indicates, Theodore Burley's, that he saw four guys walking down St. Mary's and says, yes, that looks like them, and then Aaron Funches, in his statement, indicates I was going to turn down the street, but I don't want to do that because they might recognized me or my car. Why didn't he want to be recognized or didn't want his car to be recognized? So, the car goes down the street other than St. Mary's because the four boys who were on St. Mary's never saw the car again. At that point the four boys walking down the street see Melvin Banks come up the street.

* * *

Something happened at that school. These boys were together in that white Tempo and Mr. Banks indicates — strike that. Mr. Burley indicates in his statement that he sees four guys walking down St. Mary's and says, yes, that looks like them. He might as well have took out a death warrant and signed his name to it. If it wasn't for Mr. Burley pointing out who was them, who were the guys who were in the dispute, then we wouldn't even be here.

* * *

Well, what does Mr. Funches do? Okay. There's an argument. There's a dispute at the school. Well, I'm going to give my buddies a ride home and he does. He gives Jerry Pringle a ride home and he gives Marty Williams a ride home, but not Ted. Ted stays in the car. Ted, the guy who's in the middle of all this, and not Melvin, because Melvin is, what, I don't know, bad or what. I don't know. But Melvin doesn't get a ride home and Mr. Funches indicates as he's driving that he says I was going to turn down the street, but I didn't want to. I didn't want those guys to recognize me or my car.

* * *

I think the evidence clearly shows Leonard Ingram died for no reason. He died. The other boys were shot at. Melvin Banks did those acts and they wouldn't have happened, couldn't have happened without the assistance of his buddies, Mr. Burley and Mr. Funches. Couldn't have happened.

The prosecutor drew a vivid picture for the jury: There were three persons in the car — Burley, Funches, and the defendant. Burley and Funches admitted they were there, but it was the third person in the car who did the shooting.

The closing argument of Aaron Funches' counsel was even more direct:

And it's important that he [the prosecutor] show you when because, for example, if his theory is that Mr. Banks — and he says it was Mr. Banks, but whoever he says the shooter was — premeditated this killing, planned it, decided when he got out of the car, then he did that when he got out of the car and how do you bootstrap whatever claimed assistance my client was supposed to have provided when the premeditation occurred after he got out of the vehicle?

At another point, the same counsel said:

Now, you're going to be given some lesser included offenses, voluntary manslaughter and some other things. Listen to them. Evaluate them, but what I don't want you to do is compromise. Don't go back there and play let's make a deal and say we believe Banks is the murderer so let's get him in murder one.

I don't know what Funches did. He should have stopped or something. Let's go get him for something else.

The closing argument of counsel for Theodore Burley stressed the testimony of three witnesses "who saw Melvin Banks and they point him out clearly doing the shooting," and other testimony "that the rest of the people did nothing...." Burley's counsel also responded to the prosecutor's allegation that his client had pointed out the victims:

I don't care what that out of context statement says because it doesn't say that I intended Melvin Banks to have committed a murder, a shooting, an assault, anything, and it doesn't say that I knew that Melvin Banks intended an assault, a murder, a shooting, anything and it's hard to look at the paper or it's hard to hear what someone else says and read into things, read into it, things that are not there.

It doesn't say that he was anything but a passenger in a motor vehicle which he was not operating, which he had no control over and he certainly had no control over Melvin Banks.

The explicit accusations made against the defendant in these closing arguments leave no doubt that the redaction was totally ineffective in buffering the defendant from the prejudicial effects of the admitted statements. Under the circumstances, "the risk that the jury will not, or cannot, follow instructions is so great, and the consequences of failure so vital to the defendant, that the practical and human limitations of the jury system cannot be ignored." Bruton, supra, p 135.

We conclude that the cautionary instructions given by the trial court were constitutionally inadequate. The instructions failed to insure that the jury would consider the codefendants' statements in a limited fashion — as evidence against the codefendants but not the defendant. The admission of Funches' and Burley's statements into evidence deprived the present defendant of his right of confrontation.

III

A Bruton violation does not automatically require reversal of the defendant's conviction:

In some cases the properly admitted evidence of guilt is so overwhelming, and the prejudicial effect of the codefendant's admission is so insignificant by comparison, that it is clear beyond a reasonable doubt that the improper use of the admission was harmless error. [Schneble v Florida, 405 U.S. 427, 430; 92 S. Ct. 1056; 31 L. Ed. 2d 340 (1972). See also Brown v United States, 411 U.S. 223; 93 S. Ct. 1565; 36 L. Ed. 2d 208 (1973); Harrington v California, 395 U.S. 250; 89 S. Ct. 1726; 23 L. Ed. 2d 284 (1969); People v Wright, 408 Mich. 1, 26-27, 30; 289 NW2d 1 (1980); People v Robinson, 386 Mich. 551, 563; 194 NW2d 709 (1972).]

The questionable reliability of a codefendant's statement is the theoretical underpinning of the Bruton rule. In the instant case, the statements of Funges and Burley constitute presumptively unreliable hearsay — statements of self-exoneration which are favorable, not adverse, to the penal interests of the codefendants.^[9] It was recognized, even before Bruton, that when one person accuses another of a crime under circumstances in which the declarant stands to gain by inculpating another (and exculpating himself), the accusation is presumptively suspect and therefore must be subject to cross-examination. Douglas v Alabama, supra. In Douglas, the accomplice's confession, admitted in the guise of cross-examination to refresh the testifying accomplice's recollection, was "of crucial importance" because it identified the defendant as the person who fired the gun. Id., p 417. One shot had been fired and it was certainly in the accomplice's interest to inculpate the defendant. The Court in Douglas reversed the defendant's conviction:

This case cannot be characterized as one where the prejudice in the denial of the right of cross-examination constituted a mere minor lapse. The alleged statements clearly bore on a fundamental part of the State's case against petitioner. The circumstances are therefore such that "inferences from a witness' refusal to answer added critical weight to the prosecution's case in a form not subject to cross-examination, and thus unfairly prejudiced the defendant." Namet v United States, 373 U.S. 179, 187 [83 S. Ct. 1151; 10 L. Ed. 2d 278 (1963)]. [Id., p 420.]

Similarly, in the case at bar, the codefendants' statements provided a critical element of the prosecution's case. Each claimed total innocence for himself and pointed to the defendant as the one who shot the gun. Their statements served a dual purpose — to diminish their culpability and enhance the guilt of the defendant. Unfortunately, the defendant was denied the opportunity to cross-examine them despite these prejudicial circumstances.

The evidence introduced by the prosecutor in this case, apart from the out-of-court statements of the two codefendants, included the testimony of the decedent's three companions. Lawrence Jordan, the young man who was wounded, testified that he saw the defendant pull a gun from his jacket pocket and start shooting after he said "I have a present for you all...." Jordan identified the defendant "[f]rom the clothes he was wearing."

Sean Davis, who was not shot, testified that defendant said "he had a surprise for us" and that "[w]e seen the gun coming out of his jacket pocket" and "[h]e started shooting." Larry Harris testified that he saw the defendant "walk around the corner and said I have a surprise for you and started shooting and started running."

Although this testimony without a doubt weakens the defendant's defense, it is not without its own vulnerabilities. This is not a case in which the eyewitnesses identified someone they knew well. Jordan, Davis, and Harris had never seen the defendant before the shooting incident. Jordan and Davis admitted that they had learned his name from a newspaper article and, although Jordan had picked out the defendant's photograph during a showup, he admitted that he had seen the defendant's photograph in a newspaper.

Moreover, Jordan, Davis, and Harris were all friends and members of a rival group which had allegedly been involved in prior confrontations with at least one of the codefendants. Obviously, their friendship and rivalry are factors which arguably could have influenced their testimony at trial and which provided legitimate grounds for attacking their credibility.

Although two of the three companions and several other witnesses identified the defendant as getting into a car near the high school just before the incident, the defendant admitted being in the car. He denied, however, that he went to the shooting site.

In Schneble, supra, p 432 the Court explained:

[W]e must determine on the basis of "our own reading of the record and on what seems to us to have been the probable impact ... on the minds of an average jury," Harrington v California, supra, at 254, whether Snell's admissions were sufficiently prejudicial to petitioner as to require reversal. In Bruton, the Court pointed out that "[a] defendant is entitled to a fair trial but not a perfect one." 391 U.S. at 135 quoting Lutwak v United States, 344 U.S. 604, 619 [73 S. Ct. 481; 97 L. Ed. 593] (1953). Thus, unless there is a reasonable possibility that the improperly admitted evidence contributed to the conviction, reversal is not required. See Chapman v California, 386 U.S. 18, 24 [87 S. Ct. 824; 17 L. Ed. 2d 705] (1967).

In the instant case, we conclude that the "minds of an average jury" would have found the prosecution's case "significantly less persuasive" had the statements of the codefendants been excluded. Schneble, supra, p 432. There is a very real possibility that the improperly admitted statements swayed the jury. The testimony of the three companions, while damaging to the defendant, would have born considerably less weight in the context of the defendant's defense of misidentification, without the accusations of defendant by Funches and Burley. Under these circumstances, we conclude that the error in admitting the codefendants' statements was not harmless.

We reverse the decision of the Court of Appeals and remand to the trial court for a new trial.

LEVIN, BRICKLEY, and MALLETT, JJ., concurred with GRIFFIN, J.

CAVANAGH, C.J. (concurring).

I agree wholeheartedly with the result reached by my Brother GRIFFIN, and with his careful analysis, as far as it goes. That analysis fully supports and demonstrates the wisdom of Justice GRIFFIN'S conclusion that the so-called "redaction" applied to the incriminating codefendant confessions in this case was wholly inadequate to protect defendant Banks against the danger that the jury, in violation of Banks' right of confrontation, would improperly rely on the unreliable hearsay statements in those confessions as evidence against him. However, because I believe the factors identified by Justice GRIFFIN which require finding the redaction in this case inadequate will inevitably arise in every case where the redaction is of a similarly limited and superficial nature, I would prefer to go further and enunciate a categorical rule that this sort of redaction simply does not and cannot suffice to vindicate the right of confrontation and the rule against hearsay.

As Justice GRIFFIN discusses, the leading case regarding the redaction of codefendant confessions is Richardson v Marsh, 481 U.S. 200; 107 S. Ct. 1702; 95 L. Ed. 2d 176 (1987). The codefendant confession in Richardson was redacted to remove all references, not only to the complaining defendant's name, but also to his activities and his very existence. The majority in Richardson held that redaction of that type — which I shall call "complete" redaction — is categorically sufficient to satisfy Confrontation Clause concerns. As Justice Stevens pointed out in dissent, however, even a completely redacted codefendant confession, in conjunction with other evidence at trial, may permit the jury to draw conclusions on the basis of the hearsay confession which are devastatingly prejudicial to the complaining defendant. See id. at 213-216. I would be inclined to follow Justice Stevens' reasoning in interpreting Michigan's Confrontation Clause, see Const 1963, art 1, § 20. I need not explore that issue further, however, because even accepting arguendo the majority's position in Richardson, this case involves a dramatically different kind of redaction.

In this case, the crucial hearsay confessions of Funches and Burley were each "redacted" only in the sense that Banks' name and either Funches' or Burley's name (depending on whose confession it was) were replaced by the term "blank." The detailed descriptions of Banks' alleged criminal conduct remained in both confessions as they were presented to the jury. While my Brother GRIFFIN persuasively demonstrates how easy it must have been for the jury in this case to fill in those "blanks" and identify Banks in the confessions, he goes no further, as a general matter, than to conclude that Richardson's "rule of admissibility per se is simply not appropriate for the form of redaction used in the case at bar," ante, p 420, and to find the redaction inadequate in this case on the basis of an individualized review of the evidence and argument presented at trial. While I agree that the evidence and argument presented at trial in this case make it especially obvious that the jury must have seen through the superficial redaction used here, I think the particular facts of this case merely exemplify the categorical insufficiency of such superficial, "name-only" redaction. For reasons set forth below, I find it impossible to conceive of any joint-trial, codefendant-confession situation where such superficial redaction could ever suffice to protect the defendant's rights, and I would therefore enunciate a categorical rule of exclusion of such superficially redacted confessions.

First, it should be noted that there is a fundamental difference between complete and name-only redaction. While it was assumed for purposes of analysis in Richardson that the codefendant confession was substantively inadmissible against the complaining defendant, complete redaction of the kind employed in Richardson necessarily removes from such a confession all statements directly accusing the complaining defendant. The statements left over following such complete redaction will be statements that might, therefore, have a much higher likelihood of qualifying for substantive admission under, for example, the "statement against interest" hearsay exception. See MRE 804(b)(3); FRE 804(b)(3).^[1] In dramatic contrast, name-only redaction necessarily leaves in place for the jury's consideration all the presumptively unreliable accusatory details relating to the nonconfessing defendant, with only the removal of his actual name providing a dubious shield of protection.

Second, the very fact that the mysterious "blank" has been substituted for the defendant's name, coupled with the required cautionary instruction, will inevitably arouse the jury's suspicions and entice it to draw the very inferences sought to be precluded. By contrast, the very point of complete redaction is to successfully hide the fact that the confession ever referred to the other defendant. I do not think we can safely assume that modern-day jurors, steeped in television legal dramas, are without sophistication in these matters. When jurors — presented with a joint trial of several codefendants and a confession by one codefendant which refers to one or more accomplices identified only as "blank" — are pointedly instructed not to consider that confession as evidence against the accomplices on trial with the confessor, it would be almost an insult to their intelligence to suppose that they would not deduce that the "blanks" are indeed the other defendants.^[2] After all, why would there be any need for a cautionary instruction if the confession did not contain potentially incriminating evidence regarding the other defendants?^[3]

Finally, I note that Justice GRIFFIN places special emphasis on the fact that the prosecutor's closing argument tended to erode the effectiveness of the redaction. See ante, pp 423-427. I fully agree with my colleague's analysis in this regard, but I would note that this tendency will inevitably exist in every joint trial with one or more superficially redacted codefendant confessions, no matter how conscientious the prosecutor is. The prosecutor's job is to obtain convictions of all the defendants being jointly tried, and he will properly try to do so by relying on whatever available and admissible evidence tends to show that the nonconfessing defendant participated in the joint criminal scheme and was present at the key stages of the transaction. Even where the prosecutor scrupulously avoids mentioning the redacted codefendant confession with regard to the nonconfessing defendant, the case he presents — assuming it is strong enough to convict — will necessarily draw a picture for the jury which will squarely frame the nonconfessing defendant as the "blank" in the redacted confession. I do not see how a prosecutor properly and vigorously constructing a case against a defendant like Banks could ever avoid giving the jurors (even unintentionally) precisely the hints and clues they would need to decipher the code of a superficially redacted confession.

Justice GRIFFIN'S analysis is a model for trial judges to follow. A sensitive and proper application of that analysis should lead to the rejection, in every case, of the adequacy of the superficial kind of redaction employed here. I believe, however, that Justice GRIFFIN'S analysis — which closely follows Justice Stevens' approach in his dissenting opinion in Richardson, see ante, pp 419-420 — is best reserved for those cases, like Richardson itself, which involve complete redaction. For the reasons stated above, in the interest both of better protecting the constitutional rights of defendants and of providing better and more certain guidance to the prosecutors and trial judges of this state, I would adopt today a categorical rule excluding superficial, name-only redactions.

RILEY, J. (dissenting).

I do not doubt the sincerity to which the majority attaches the guarantees of the Confrontation Clause with the facts involved in the instant case. Nor do I reject the notion that the right to confront one's accusers is one of the most important, if not the most important, aspect of a fair trial. However, I firmly believe that the majority errs in its analysis of Richardson v Marsh, 481 U.S. 200; 107 S. Ct. 1702; 95 L. Ed. 2d 176 (1987), and therefore reached the erroneous conclusion that defendant's confrontation rights were violated. In my opinion, the statements of codefendants Funches and Burley were not admitted "against" defendant, and therefore his confrontation rights were not violated. I respectfully dissent.

I

In Bruton v United States, 391 U.S. 123; 88 S. Ct. 1620; 20 L. Ed. 2d 476 (1968), the United States Supreme Court held that the defendant was deprived of his confrontation rights under the Sixth Amendment where, in a joint criminal trial, a nontestifying codefendant's confession named the accused as a participant in a robbery. In assuming that the codefendant's implicating confession was inadmissible against the accused, the Bruton Court centered upon the effectiveness of a limiting instruction and held that it was constitutionally insufficient to support Bruton's conviction.

Two decades later in Richardson v Marsh, the Court considered another case where a codefendant's confession was introduced as an admission against party interest in the accused's joint criminal trial. However, the prosecutor redacted the name and all references to the accused's participation in the criminal episode in the codefendant's statement. The trial judge admitted the codefendant's confession and instructed the jury, which ultimately convicted the accused under an aiding and abetting theory, to only consider the statement against the declarant. After her conviction was affirmed on appeal, and being denied habeas corpus relief in federal district court, the accused turned to the Sixth Circuit Court of Appeals. Richardson argued that the introduction of the codefendant's confession violated her confrontation rights because the jury could infer from the evidence admitted at trial that she was the person to whom the codefendant referred to in his statement. Marsh v Richardson, 781 F2d 1201, 1213 (CA 6, 1986). The Court of Appeals agreed with her and held that Bruton requires a determination of whether there exists a "substantial risk that the jury would consider [a codefendant's] statement in" light of all the other evidence admitted at trial. Id.

The United States Supreme Court granted certiorari, reversed, and expressly rejected the Sixth Circuit's "contextual implication" analysis. Instead, the Court adopted a "facially incriminating" standard to test confrontation challenges. 481 U.S. 208. The Court of Appeals for the Second Circuit has best summarized the "facially incriminating" test as follows:

"[D]efendant's Bruton rights [are] violated ... only if the statement, standing alone, would clearly inculpate him without introduction of further independent evidence." United States v Wilkinson, 754 F2d 1427, 1435 (CA 2, [1985]), [as reaffirmed] in Richardson v Marsh, 481 U.S. 200; 107 S. Ct. 1702; 95 L. Ed. 2d 176 (1987). [United States v Tutino, 883 F2d 1125, 1135 (CA 2, 1989).]

The Richardson decision makes it clear that the distinction between the facts of that case, and those of Bruton, control the decision whether a codefendant's implicating statement is constitutionally acceptable.

In Bruton, the codefendant's confession "expressly implicat[ed]" the defendant as his accomplice. Thus, at the time that confession was introduced there was not the slightest doubt that it would prove "powerfully incriminating." By contrast, in this case the confession was not incriminating on its face, and became so only when linked with evidence introduced later at trial (the defendant's own testimony). [481 U.S. 208. Citations omitted.]

In the instant case, the majority has adopted a Sixth Amendment confrontation analysis that is directly at odds with the holding of Richardson v Marsh. The majority assumes that the "facially incriminating" test is too difficult to apply because, in cases such as this one, a jury will be able to "fill in [the] blank[s]." Thus, under the majority's proposed rule, a trial court would be required to determine, at the end of every joint criminal trial, where a codefendant's statement is introduced "[i]f a ``substantial risk' exists that the jury, despite cautionary instructions, will consider a codefendant's out-of-court statement in deciding the defendant's guilt...." Ante, p 421.

In my opinion, the majority's adoption of the "contextual implication" test is not only more difficult to apply than the "facially incriminating" test, but it defies the guarantees of the Confrontation Clause and clogs the wheels of criminal justice. Indeed, Justice Scalia offered two persuasive reasons why the "contextual implication" test should not be used in confrontation challenges. He first noted that

[t]he "contextual implication" doctrine articulated by the Court of Appeals would presumably require the trial judge to assess at the end of each trial whether, in light of all of the evidence, a nontestifying codefendant's confession has been so "powerfully incriminating" that a new, separate trial is required for the defendant. This obviously lends itself to manipulation by the defense — and even without manipulation will result in numerous mistrials and appeals. [481 U.S. 209.]

Justice Scalia's observation is well taken. Any resourceful defense attorney would be able to manipulate the trial process by admitting evidence that possibly links a client with the omitted reference in the codefendant's statement, thus defeating the purpose of having joint criminal trials and placing additional burdens on an already beleaguered criminal justice system.

Second, the majority implicitly assumes that a jury neither can comprehend nor follow a limiting instruction given by a trial judge to consider a statement only against the declarant, and not against the accused. Here again, Justice Scalia explained why this fear is misguided:

Where the necessity of such linkage is involved, it is a less valid generalization that the jury will not likely obey the instruction to disregard the evidence. Specific testimony that "the defendant helped me commit the crime" is more vivid than inferential incrimination, and hence more difficult to thrust out of mind. Moreover, with regard to such an explicit statement the only issue is, plain and simply, whether the jury can possibly be expected to forget it in assessing the defendant's guilt; whereas with regard to inferential incrimination the judge's instruction may well be successful in dissuading the jury from entering onto the path of inference in the first place, so that there is no incrimination to forget. In short, while it may not always be simple for the members of a jury to obey the instruction that they disregard an incriminating inference, there does not exist the overwhelming probability of their inability to do so that is the foundation of Bruton's exception to the general rule. [481 U.S. 208.]

The assumption that jurors are able to follow a trial judge's instructions fully applies when rights guaranteed by the Confrontation Clause are at issue. Tennessee v Street, 471 U.S. 409, 415, n 6; 105 S. Ct. 2078; 85 L. Ed. 2d 425 (1985) (quoting Frazier v Cupp, 394 U.S. 731, 735; 89 S. Ct. 1420; 22 L. Ed. 2d 684 [1969]). The question in this case, as in Bruton, therefore becomes whether we may rely on this assumption that the jurors followed the instructions given them by the trial judge. In my opinion, the record supports the presumption that the jurors in the instant case followed the trial judge's instructions and convicted the defendant on the basis of the evidence admitted at trial, without the benefit of the codefendants' statements.

It is evident that the trial judge was very concerned with selecting an impartial and fair jury. He first had all of the witnesses involved sequestered from the courtroom, and then allowed the attorneys to conduct voir dire, which spanned three days, and dismissed just about every person who did not want to be on the jury. The prosecutor, in conducting his voir dire, was the first person to ask the jurors if they could consider a statement individually with regard to each defendant, and repeated the theme two other times. Codefendant Funches' attorney also asked the question once, and Banks' defense counsel asked the question five times. Upon their selection as jurors, they were sworn under oath to uphold the law given them.

Before the trial commenced, the trial judge made sure the witnesses were sequestered and instructed the jury to consider the codefendants' statements only against the declarant. Thus, even before this trial began, the jury was instructed ten times to consider a statement individually with regard to each defendant.

The prosecutor again emphasized the limited nature of the statements in his opening remarks. He informed the jury that he would be introducing statements made by both codefendants Burley and Funches, but again cautioned the jury to only consider the use of a statement against each declarant. Before the introduction of Burley's statement, the prosecutor asked the trial judge to instruct the jury to consider the statement only against the declarant, which he did, and the same was done when Funches' statement was introduced. During closing arguments, the prosecutor again cautioned the jury to consider the evidence individually with regard to each defendant, and then went through all the evidence, not just the statements, and concluded that it was defendant who did the shooting and Burley and Funches who participated as aiders and abettors.^[1] After the defense closed and the prosecutor completed his rebuttal, the trial court instructed the jury, for a fourth time, about how it should use the statements.

Under these circumstances, there is no sound reason why the majority should apply the Bruton exception rather than the rule. The undeniable fact of the matter is that the acquittal of both Funches and Burley is proof positive that the jury did follow its instructions and did compartmentalize the evidence regarding each of the defendants. So too, there exists no "overwhelming probability" that the jury disregarded its limiting instructions when deliberating on defendant's case, and therefore, he was not denied the opportunity to confront the witnesses "against him."

Can it be that a majority of this Court no longer believes that the cornerstone of our adversarial trial system is the notion that a jury will follow the instructions given to them? I would presume that jurors follow the court's instructions, and, in this case, I believe they did.

II

In addressing the issue the Court decides today, I would utilize the "facially incriminating" confrontation analysis adopted in Richardson v Marsh. That is, in determining if defendant's Bruton rights are violated, examine the declarant's statement, standing alone, and decide whether it clearly inculpates the accused without the introduction of other properly admitted evidence. Richardson requires nothing less.

The statements of Funches and Burley were admitted as admissions against a party-opponent under MRE 801(d)(2), and not "against" the defendant.^[2] To insure that the jury would follow the trial judge's limiting instructions, the statements were redacted to exclude all direct references to defendant by name and description. There is no dispute that the redacted statements were not "powerfully incriminating" of defendant on their face.

To illustrate, if the jury had been sent to deliberate on defendant's case after having heard only the redacted statements of Funches or Burley, or both for that matter, it could have only arrived at a verdict of not guilty. Obviously then, some meaning had to be given to each respective statement before it could have had meaning and become incriminatory of defendant. I believe that where such "linkage" is involved, there does not exist an "overwhelming probability" that a jury will not be able to disregard incriminating inferences. This is the foundation of Bruton's narrow exception.

Federal courts have not encountered any difficulty in applying the "facially incriminating" test in Sixth Amendment confrontation challenges. Even before Richardson was decided, most of the federal appellate courts held that a statement, made by one defendant that is not inculpatory of a codefendant on its face, is admissible in a joint criminal trial (when the trial judge instructs the jury to consider the statement only against the declarant), even though other evidence in the case indicates that a codefendant not mentioned in the statement was also involved in the activities described. See, generally, United States ex rel Cole v Lane, 752 F2d 1210, 1216 (CA 7, 1985); United States v Belle, 593 F2d 487, 493 (CA 3, 1979); United States v Stewart, 579 F2d 356 (CA 5, 1978); United States v Brown, 551 F2d 639, 647 (CA 5, 1977), rev'd on other grounds 569 F2d 236 (CA 5, 1978); United States v Dady, 536 F2d 675 (CA 6, 1976); United States v Hicks, 524 F2d 1001 (CA 5, 1975); United States v Trudo, 449 F2d 649, 652-653 (CA 2, 1971); Slawek v United States, 413 F2d 957, 960-964 (CA 8, 1969).

Moreover, after Richardson was decided, the federal appellate courts have applied the "facially incriminating" test with relative ease. See, for example, United States v Kendrick, 853 F2d 492, 496, n 3 (CA 6, 1988); United States v Coppola, 788 F2d 303 (CA 5, 1986); United States v Burke, 700 F2d 70, 85 (CA 2, 1983).

The facts in the instant case fit into one category not addressed in Richardson, that is, when the names and references to a codefendant are redacted and replaced by neutral pronouns, with no indication to the jury that the original statement contained the codefendants' names, and where the statement, standing alone, does not otherwise connect the codefendant to the crimes.

In the leading federal circuits, the general rule afforded in these cases is that a redacted statement in which the names of codefendants are replaced by neutral pronouns, with no indication to the jury that the original statement contained actual names, and where the statement, standing alone does not otherwise connect the codefendants to the crimes, may be admitted without violating a codefendant's Bruton rights. See, generally, United States v Tutino, supra, 1135 (names and references were replaced by the words "others," "other people," and "another person"); United States v Briscoe, 896 F2d 1476, 1501-1502 (CA 7, 1990) (names and references replaced with the word "we"); United States v Alvarado, 882 F2d 645, 652 (CA 2, 1989) (name and reference replaced with the words "another person"); United States v Benitez, 920 F2d 1080, 1087 (CA 2, 1990) (name and reference replaced by the word "friend").

I believe that Funches' and Burley's statements were sufficiently redacted to omit the names and references to the defendant. Their statements, standing alone, do not implicate defendant in the shooting death of Leonard Ingram and the injury of Larry Jordan. I would hold that the guarantees of confrontation have been met, and therefore, defendant's Bruton challenge has failed.

III

I also believe that the majority erred in its harmless error analysis. In my opinion, the evidence overwhelmingly supports the jury verdict. The evidence brought out at trial, notwithstanding Funches' and Burley's statements, clearly indicate that defendant shot and killed Ingram and injured Jordan.

Larry Jordan testified that he, along with Sean Davis and Larry Harris, were with Leonard Ingram when defendant shot him. Jordan put defendant in the car driven by codefendant Burley and described the first time he saw defendant that day. He also described the clothes defendant wore and the words defendant used before he killed Ingram. Jordan also testified that defendant shot him in the leg and that he identified defendant in a photo identification before defendant's arrest. This testimony is consistent with the testimony of the other two eyewitnesses to the shooting, Sean Davis and Larry Harris.^[3]

Tony Ulmer, a thirteen-year-old neighborhood boy, testified that he was sitting in his house when he heard four shots being fired. He told the jury that he did not see the shooter's face, but he did see him running away and described the clothing that he wore. The clothing matched the description given by Jordan of the items defendant wore on the day in question. Jerry Pringle testified that he was in the car with Funches and Burley, as well as Marty William. He also testified that he was the last person dropped off, leaving defendant, Funches, and Burley remaining. Dawn Brookins, who was Burley's girl friend at the time of the shooting, testified that defendant was in the car on the day in question. Bryant Berry, the person who was probably the target of the shooting, testified that he saw defendant in Burley's car and that he was later kicked out of high school for carrying a gun.

Defendant presented his witnesses and an alibi defense. He testified that he was not wearing the clothes described by the eyewitnesses and was not in the vicinity of the shooting at the time it occurred, and that the people who placed him in Burley's car were mistaken.

In sum, three witnesses testified that they saw defendant kill Ingram and shoot Jordan. One testified that he saw defendant running from the scene. Five witnesses placed defendant in Burley's car at the time of the shooting. Defendant's alibi defense was so weak that a reasonable jury would reach the same conclusion as the one in the instant case and disbelieve him. Given these overwhelming facts, I would apply the harmless error rule and affirm the defendant's conviction.

IV. CONCLUSION

The majority's "contextual implication" analysis disregards the fact that jurors are law-abiding men and women who consciously strive to uphold and apply the law as given. The meticulous jury selection process here, which spanned three days, insured that those jurors selected were most capable of faithfully following the law given them and able to compartmentalize the evidence with regard to each defendant. If we are to follow the rule that jurors are presumed to follow their instructions, I believe that after being told fifteen times that they are to consider the statements individually, they will — and in this case, they did.

BOYLE, J., concurred with RILEY, J.

NOTES

[1] Defendant made a timely motion for a separate trial. Separate preliminary examinations were held, but the prosecutor successfully sought to consolidate the cases for trial, over defense objection. The court thereafter refused to vacate the order of consolidation, and another judge denied a subsequent motion for a separate trial. The Court of Appeals declined interlocutory review.

[2] A redacted statement is one from which, minimally, direct reference to the defendant is deleted.

[3] The parties also agreed to excise any description of defendant Banks from the statements and to delete certain hearsay evidence contained within the statements.

[4] Funches and Burley faced charges identical to those of the present defendant: first-degree murder, MCL 750.316; MSA 28.548, possession of a firearm in the commission of a felony, MCL 750.227b; MSA 28.424(2), and assault with intent to commit murder, MCL 750.83; MSA 28.278.

[5] The Court further ordered that the instant case be argued and submitted together with People v Watkins (Docket No. 86776), People v Hunter (Docket No. 86806), and People v Phillips (Docket No. 87091).

[6] every criminal prosecution, the accused shall have the right ... to be confronted with the witnesses against him ... [Const 1963, art 1, § 20.]

[7] On the same day that Richardson was decided, the Court decided Cruz v New York, 481 U.S. 186; 107 S. Ct. 1714; 95 L. Ed. 2d 162 (1987), which dealt with an alleged Bruton error involving "interlocking" confessions — where the defendant has made a full, voluntary confession which is nearly identical to the confession of his codefendant. The Cruz Court held that, although introduction of the defendant's own interlocking confession cannot cure the Confrontation Clause violation caused by the introduction of a nontestifying codefendant's confession, it might, in some instances, render that violation harmless. Since the defendant in the instant case did not confess, we do not address the distinctive concerns which arise in interlocking confession cases. See id., pp 191-193.

[8] Whether or not the codefendant's statement is crucial to the prosecution's case — i.e., has a "devastating" effect on the defendant — bears on the harmlessness of a Bruton violation, not the existence of such an error. Cruz v New York, n 7 supra, pp 191-192.

[9] In the instant case, as in Bruton, supra, p 128, n 3, the inadmissibility of the codefendants' statements against the defendant is not contested. See also Lee v Illinois, 476 U.S. 530, 542; 106 S. Ct. 2056; 90 L. Ed. 2d 514 (1986):

[W]e are not here concerned with the effectiveness of limiting instructions in preventing spill-over prejudice to a defendant when his codefendant's confession is admitted against the codefendant at a joint trial. Rather ... [h]ere ... the State sought to use hearsay evidence as substantive evidence against the accused.

[1] For example, the inculpatory effect of the redacted confession in Richardson, as Justice Stevens explained, was that the confessor described having a conversation with a third accomplice while driving to the murder victims' home, in which he explicitly planned their deaths. While the confession was carefully redacted to remove any reference to the fact that the complaining defendant was allegedly present in the car at the time, a different witness testified at trial that the defendant arrived at the house in the same car with the confessor and the other accomplice. See 481 U.S. 215. Thus, the hearsay confession, in conjunction with the separate evidence, provided crucial proof that the defendant must have known about the plan to kill the victims. The jury had only to add two and two together. And yet (without expressing any firm view on the merits), it is conceivable that the confessor's hearsay statement, shorn of any added accusatory reference to the presence of the complaining defendant, might have qualified for substantive admission as a statement against penal interest. The separate and crucial allegation that the defendant was present and heard the plans for the murder would certainly have been an intrinsically suspect and unreliable "spread-the-blame" accusatory hearsay statement had it derived from the confession itself. But as it happened, that crucial allegation came, not from the hearsay confessor, but from a separate witness who testified at trial subject to full confrontation and cross-examination.

[2] In cases (unlike the instant case) where the confessions and other evidence do not dovetail neatly enough to provide a clear indication of which "blank" is which, the danger inevitably arises that the jury will draw speculative conclusions which are not only improper, but mistaken and misguided as well.

[3] Conversely, in cases where the redaction is genuinely complete and effective, I would imagine that defense counsel might often prefer to have no cautionary instruction at all, so as to avoid unnecessarily arousing the jury's suspicions. There would be no need for a cautionary instruction if the confession were redacted so thoroughly as to make it genuinely impossible for the jury to draw any incriminating conclusions therefrom about the nonconfessing defendant.

[1] In Richardson, the prosecutor sought to undo the effect of the limiting instruction, in his closing argument, by urging the jury to use the codefendant's confession in evaluating the respondent's case. 481 U.S. 211. The Court remanded the case to the Court of Appeals to consider whether, in light of the respondent's failure to object to the prosecutor's comments, the error could serve as the basis for granting habeas corpus. Id.

In the instant case, the majority assumes that the prosecutor also urged the jury to evaluate defendant's case in light of Funches' and Burley's statements, "thereby effectively undoing the trial court's cautionary instructions." Ante, p 424. This assumption is misplaced. The quotation that the majority has taken out of the prosecutor's closing argument was his theory of what happened on the day in question. He never urged the jury to consider the codefendants' statements against defendant, nor did he undo the effect of the trial court's limiting instructions.

[2] The statements of Funches and Burley cannot be categorized as a declaration against penal interest, because the statements were not against the declarant's penal interest when made.

[3] It is important to note that all three witnesses put defendant in the car driven by the codefendant, described the clothing that defendant wore on the day of the shooting, identified defendant as the shooter, and repeated the words spoken by defendant before he killed Ingram and injured Jordan.

There was no testimony elicited at trial that Ingram, Jordan, Harris, or Davis were part of the same "rival gang" as the majority has assumed. See ante, pp 429-430.