Carter v. Rafferty, 621 F. Supp. 533 (D.N.J. 1985)
November 13, 1985
John J. RAFFERTY, Superintendent, Rahway State Prison, and Irwin I. Kimmelman, The Attorney General of the State of New Jersey, Respondents.
John ARTIS, Petitioner,
Christopher DEITZ, Chairman, Parole Board of the State of New Jersey and Irwin I. Kimmelman, the Attorney General of the State of New Jersey, Respondents.
United States District Court, D. New Jersey.
*534 Leon Friedman, Hofstra University Law School, Hempstead, Grover G. Hankins, Charles E. Carter, N.A.A.C.P., Brooklyn, N.Y., Harold J. Cassidy, Cassidy Despo & Foss, Red Bank, N.J., for petitioners Carter and Artis.
Myron Beldock, Edward S. Graves, Beldock Levine & Hoffman, New York City, Ronald J. Busch, Busch, Busch & Busch, New Brunswick, N.J., for petitioner Carter.
Lewis M. Steel, Steel & Bellman, P.C., New York City, Louis Raveson, Rutgers School of Law, Newark, N.J., for petitioner Artis.
John P. Goceljak, First Asst. Pros., Joseph A. Falcone, Passaic County Prosecutor, Paterson, N.J., for respondent.
SAROKIN, District Judge.
The court is called upon in this matter to issue the "Great Writ," long recognized as "the best and only sufficient defense of personal freedom." The petitioners, convicted of a triple murder that took place nearly 20 years ago, continue to protest their innocence and allege that serious violations of their constitutional rights occurred at trial. The court's responsibility to examine these allegations is in no way diminished by the existence of extensive review by the learned state courts which have ruled in this case. Indeed, it is tempting to presume the correctness of those rulings, but this court is charged to resist such temptation lest it fail in its duty to independently analyze the constitutional violations asserted in the petitions for habeas corpus relief. In so doing, the court has determined that the writ must issue. The extensive record clearly demonstrates that petitioner's convictions were predicated upon an appeal to racism rather than reason, and concealment rather than disclosure.
The jury was permitted to draw inferences of guilt based solely upon the race of the petitioners, but yet was denied information which may have supported their claims of innocence. To permit convictions to stand which have as their foundation appeals to racial prejudice and the withholding of evidence critical to the defense, is to commit a violation of the Constitution as heinous as the crimes for which these petitioners were tried and convicted.
Were it not for these grave constitutional violations, the court concludes, for the reasons hereafter set forth, the guilty verdicts of the jury might well have been otherwise.
At approximately 2:30 a.m. on June 17, 1966, two armed black men entered the Lafayette Bar & Grill in Paterson, New *535 Jersey and opened fire. The bartender, James Oliver and a patron, Fred Nauyaks, were killed immediately. A second patron, Hazel Tanis, died one month later from her wounds and a third, William Marins, was partially blinded after being shot in the head.
Arrested four months later for the murders were Rubin Carter, a well-known professional boxer who lived in Paterson, and who was, at 30 years old, reaching the peak of his career, a contender for the middleweight crown; and 20 year-old John Artis, who was about to enter college on a scholarship.
As the case took a tortuous and often circuitous route through the New Jersey courts, the circumstances surrounding the killings and subsequent prosecution of Carter and Artis have become a mosaic. The picture of the evidence painted by petitioners and respondents is often conflicting and sometimes exceptionally murky. The accounts of many important witnesses, especially that of Alfred Bello, the only "eyewitness" to testify at the 1976 trial, have changed; some witnesses have died; the memories of those who survive have grown hazy. But from thousands of pages of testimony spanning two trials and numerous hearings, the parties have reconstructed two drastically different versions of the events that tragic night. The conflicting evidence is reviewed below (See The Brady violation) but a brief summary of the evidence introduced at the second trial is presented here.
Patricia Valentine lived above the tavern and was awakened by gunshots about 2:30 a.m. She first ran to her window and saw two men leave the scene in a white car, then ran downstairs to the bar. Alfred Bello and Arthur Bradley were in the process of breaking into a nearby factory. Bello, who was standing lookout, was either in or outside of the bar (a main point of contention). Within minutes, police arrived at the scene and took statements from Bello, Marins and Valentine. A description of the car was sent out on police radio. A few minutes later, a Paterson police officer who just given up a brief chase of two cars, one white and one black, speeding out of town, started driving back into Paterson. He stopped a white car leased by Carter about 14 blocks away from the Lafayette Bar. Artis was driving, John Royster was sitting in the front seat, and Carter was alone in the back seat. The car was not speeding and there were no weapons in sight. Carter told the officer that the men were driving to his home, about six blocks away, to obtain money, and the car was allowed to go. Fifteen minutes later, Carter's car was observed outside the La Petite Bar about 10 blocks west of the Lafayette. About five minutes later, the car was sighted for a third time, with only Carter and Artis in the vehicle. This time the police escorted the car and its occupants to the crime scene.
From the evidence, it appears neither Bello nor Valentine identified the petitioners at the scene, and there is considerable dispute as to the identification of the car at the scene and again later that evening at the police station. The petitioners were taken to the station and to the hospital where the two survivors did not identify them. The petitioners were questioned, given polygraph examinations and released about 7 p.m. on June 17.
Meanwhile, police searched the car in which they later alleged that they found a live .12 gauge shotgun shell in the trunk and a live .32 caliber shell on the floor of the front seat. There was considerable dispute about this evidence.
The petitioners voluntarily testified before a Passaic County Grand Jury which did not return any indictments. In July and October, Bello told a Paterson police officer that he had seen Carter and Artis at the crime scene. Bello and Bradley identified Carter as one of the two black men they saw coming from the Lafayette Bar with weapons in their hands. Bello, but not Bradley, identified Artis as the second man. On the basis of these identifications, the petitioners were indicted and ultimately convicted of murder in June of 1967.
*536 In late 1973, Fred Hogan, a state Public Defender's Office investigator assigned to Carter's case began contacting Bello and Bradley about the case. Hogan had communicated with Richard Solomon, a documentary film-maker, Selwyn Raab, who then worked for a New York television station, and Hal Levinson, who worked for Raab at the station. Raab and Levinson eventually talked with Bello about recanting his testimony. After Bradley gave Hogan a statement recanting his identification of Carter, Hogan continued to talk with Bello, who in a written statement from jail in September 1974 first recanted his 1966 testimony (37T125). The recantation resulted in a hearing and began a series of events that ultimately set the stage for the second trial in 1976.
At the retrial, there was considerable testimony concerning alleged inducements made to Bello by both the prosecution and the defense camp in order to change his story. (See: The Brady Violation, infra) There was also evidence presented, and disputed by the defense, that Carter attempted to create a false alibi for the 1967 trial. (See: The Brady Violation, infra) The petitioners were both found guilty of first degree murder. Carter was sentenced to two consecutive and one concurrent life sentences and remains in prison. Artis, who received a lesser sentence, was released on parole for a ten year term beginning December 22, 1981.
The petitioners were originally tried and convicted in 1967 of three counts of first degree murder. Life sentences were imposed after a jury recommendation of mercy on June 29, 1967. The convictions were affirmed on direct appeal to the New Jersey Supreme Court. State v. Carter, 54 N.J. 436, 255 A.2d 746 (1969) ("Carter I"), cert. denied, 397 U.S. 948, 90 S. Ct. 969, 25 L. Ed. 2d 130 (1970).
On October 1, 1974, the petitioners filed a new trial motion based on the statements of the state's two key witnesses recanting their 1967 identification testimony. The original trial judge denied the motion. State v. Carter, 136 N.J.Super. 271, 345 A.2d 808 (Cty.Ct.1974). A second new trial motion was made on January 30, 1975 and that motion was also denied. State v. Carter, 136 N.J.Super. 596, 347 A.2d 383 (Cty. Ct.1975). That decision was appealed to the New Jersey Supreme Court, which, on March 17, 1976 overturned the convictions and ordered a new trial. State v. Carter, 69 N.J. 420, 354 A.2d 627 (1976) ("Carter II"). The Supreme Court ruled that the prosecution had withheld from the defense exculpatory evidence which demonstrated that prosecutors had offered the key identification witnesses protection and help with criminal charges pending and/or threatened against them.
The case was remanded to the trial court, where there were numerous motions and hearings. The retrial began on October 12, 1976 and concluded on December 22, 1976 when the jury returned first degree murder verdicts. The judgments were appealed to the Appellate Division of New Jersey Superior Court and the state Supreme Court. During the pendency of these appeals, there was a federal habeas corpus order for a state court hearing on allegations of juror misconduct. These hearings were held in 1979 by the second trial judge, who found against the defense on all issues.
Among the new trial motions pending before the New Jersey appellate courts was a motion seeking an evidentiary hearing to determine whether the prosecution had misrepresented the results of a 1976 pretrial polygraph test given to Alfred Bello. (See: The Brady Violation, infra) The Appellate Division affirmed the convictions in an unreported opinion issued October 22, 1979. The New Jersey Supreme Court reversed, however, remanding the *537 issue to the trial court for a hearing on issues surrounding the polygraph test. State v. Carter, 85 N.J. 300, 426 A.2d 501 (1981) ("Carter III").
The remand hearing before the trial judge lasted 15 days. In an 80-page unreported opinion dated August 28, 1981 ("Opinion on Remand"), the judge found against the defense on all issues. One year later, following further briefing and oral argument, the New Jersey Supreme Court affirmed the convictions by a 4-3 majority. State v. Carter, 91 N.J. 86, 449 A.2d 1280 (1982) ("Carter IV").
Subsequently, petitioners filed another application before the trial court for an evidentiary hearing relating to allegedly exculpatory material contained in the file of a former prosecution investigator. That application was denied and appealed to the Appellate Division, which affirmed the trial court in an unreported decision dated July 2, 1985.
THE HABEAS CORPUS PETITIONS
The instant petitions for habeas corpus were filed February 13, 1985 by Carter and February 28, 1985 by Artis. The actions were consolidated in an order dated May 6, 1985. A motion for summary judgment on seven of the petition's twelve grounds was filed by petitioners on May 25, 1985 and oral argument on the motion was held July 22, 1985, after which the court received further submissions from counsel.
The petitioners allege:
1. The state violated the requirements of Brady v. Maryland, 373 U.S. 83, 83 S. Ct. 1194, 10 L. Ed. 2d 215 (1963) by failing to disclose the results of a lie detector test given to the state's only "eyewitness".
2. The state violated the Equal Protection and Due Process rights of petitioners by improperly appealing to racial prejudice during the trial by claiming the killings were motivated by racial revenge.
3. The jury considered material not introduced as evidence and members of the jury had preexisting racial prejudice.
4. The prosecution misled the defense as to its theory of the case, withheld discovery, improperly cross-examined and unfairly denigrated defense witnesses.
5. The prosecution violated the Due Process rights of petitioners by exerting improper pressure on certain witnesses to support a false alibi claim.
6. The prosecution violated the Due Process rights of petitioners by using Bello as a witness despite proof of his "monumental" untrustworthiness.
7. The prosecution withheld a memorandum in violation of Brady showing how Bello was persuaded to change his mind.
8. The prosecution violated the petitioners' rights to a Speedy Trial by making last-minute presentations at trial of a new theory.
9. The evidence presented at trial did not meet the reasonable doubt standard.
10. The petitioners' Due Process rights were violated because public funds were not made available for investigative and expert services and because the trial judge did not properly exercise judicial discretion to limit cross examination regarding Carter's criminal record.
11. The petition's Due Process rights were violated by the trial judge's bias.
12. The state committed a Brady violation by failing to produce a file by a former investigator in the case.
Petitioners seek summary judgment on Grounds 1, 2, 4, 6, 7, 8 and 10.
During oral argument, counsel for Petitioner Carter represented that the complaint would be amended to exclude Ground 12, the identical issue rejected by the Appellate Division on July 2, 1985, so that there would be no dispute that he had presented only exhausted claims to his petition. Respondents agreed to the amendment of the complaint.
Petitioner Artis has not amended his petition to delete Ground 12, and argues that technical exhaustion of his state remedies is unnecessary because the corrective process in the New Jersey state courts is so *538 clearly deficient as to render futile any efforts to obtain relief. That issue is discussed below. (See: Exhaustion of State Remedies, infra).
THE RACIAL REVENGE THEORY
A. The State's Arguments
The court first turns to Ground 2, that the petitioners' Due Process rights were violated. During the 1967 trial, the defense argued that the state had not established any reason or motive for Carter and Artis to have committed the murders. Anticipating the same tactic at the 1976 retrial, the state contended the shotgun murder of James Oliver, a white bartender at the Lafayette Bar, was motivated by racial revenge. The state theorized that the other three bar patrons were shot only because they had witnessed Oliver's killing, and that there was no robbery, because their money was scattered about the bar after the rampage. (16T44, 19T174, 32T167-171).
The genesis for the state's theory began about six hours before the Lafayette Bar killings, when Leroy Holloway, the black owner of the Waltz Inn, another Paterson tavern, was shot and killed by Frank Conforti, a white man who had previously owned the Waltz Inn. The state introduced testimony from police officers that a large and "angry" crowd gathered outside the Waltz Inn shortly after the shooting. (32T194, 31T136-141) Within the next few hours, according to the Grand Jury testimony of Rubin Carter himself, there was talk of a "shaking" or some sort of retaliatory action among the black community. (36T153-154)
In order to tie the two killings together, the state argued that the Lafayette Bar had served primarily white patrons and was an ideal target of this anger. (31T69-71) It was located on the fringes of a black neighborhood and Oliver, its bartender, had on occasion refused to serve black customers, according to the testimony of one officer. (17T75-80) The state alleged that Carter and Artis were both aware of the murder of Holloway and the ensuing tense atmosphere in the streets, and that they were driven to action out of their friendship with Edward Rawls, Holloway's stepson and a part-time bartender at the Nite Spot, a Paterson bar frequented by Carter. (39T257-258) After he learned of the Waltz Inn shooting, Rawls took the night off from his job at the Nite Spot and went to the hospital where his stepfather had been taken, then to police headquarters, where, according to police officers' later testimony, Rawls demanded to know what the "police intended to do about the guy who killed his stepfather." One of the same officers testified that Rawls became agitated and was ordered to leave. (33T5-8) Afterwards, Rawls met Carter at the Nite Spot, where Carter extended his condolences. The state asserts that Carter then set out to search for weapons stolen from his training camp a year earlier, found the guns and committed the murders. (36T136-149).
B. Proper Use of Motive
At trial, the defense objected strenuously to the introduction of the racial revenge motive, on the grounds that such evidence would be prejudicial and that its probative value was outweighed by its prejudicial effect (31T120). The trial judge, relying on State v. Rogers, 19 N.J. 218, 116 A.2d 37 (1955), ruled that in criminal prosecutions, wherein the motive is important and material, a somewhat wider range of evidence is permitted. (31T121). In Rogers, the court allowed evidence showing loans made by a murder victim to the defendant were never repaid. Similarly, the Carter trial judge held that the proffer of motive was material and probative and had a tendency to explain conduct which would ordinarily or otherwise probably be unexplainable. (31T122). The New Jersey Supreme Court agreed, citing Rogers and its progeny, and ruled that "There is nothing inherently wrong with advancing a theory of revenge as a motive for murder, if the facts bear out the theory." Carter IV, 91 N.J. at 106, 449 A.2d 1280.
*539 It is well-established that the prosecution may introduce evidence of motive. Carter IV, at 102, 449 A.2d 1280, 1 Wigmore and Tillers, Evidence § 118, at 1697-98 (1983). "Motive cannot be shown directly, but may be inferred from facts in evidence. In the introduction of evidence to show motive, a wide range is permitted. Thus, any evidence which logically tends to show a motive, or fairly tends to explain the conduct of the accused should be permitted" 1 Wharton, Criminal Evidence, § 170 at 317-318 (13th ed. 1972). In State v. Rogers, supra, the court allowed the prosecution to introduce evidence to show the defendant had been indebted to the murder victim. The New Jersey courts have continued to follow Rogers, as have other courts throughout the country. For instance, in State v. Baldwin, 47 N.J. 379, 391, 221 A.2d 199 (1966), cert. denied, 385 U.S. 980, 87 S. Ct. 527, 17 L. Ed. 2d 442 (1966), the court allowed evidence that the defendant's victim was a prospective witness; in State v. Royster, 57 N.J. 472, 484-85, 273 A.2d 574 (1971), the court allowed evidence showing the defendant threatened the victim several weeks before the murder; in United States v. Parker, 549 F.2d 1217 (9th Cir. 1977), cert. denied, 430 U.S. 971, 97 S. Ct. 1659, 52 L. Ed. 2d 365 (1977), the court upheld the introduction of evidence of narcotics dealing as motive to commit a robbery; and in United States v. Michaels, 726 F.2d 1307 (8th Cir.1984), cert. denied, ___ U.S. ___, 105 S. Ct. 92, 83 L. Ed. 2d 38 (1984), the court allowed evidence of the car bombing death of the defendant's grandfather as motive for a subsequent bombing.
C. The New Jersey Supreme Court's Analysis
The New Jersey Supreme Court majority accepted the racial revenge theory as probative, dismissing the petitioner's claims that irrespective of its relevance, the impact of the evidence involving motive was inflammatory and improperly swayed the jury. The decision, the court said, was within the discretion of the trial judge, and absent abuse of discretion or a manifest denial of justice, should not be disturbed. Carter IV, 91 N.J. at 106-107, 449 A.2d 1280. "Ordinarily, evidence as to motive is admissible even though it may be prejudicial in the sense that it will arouse or inflame the jury." 1 Wharton, Criminal Evidence, supra, § 170, at 316. This case is, however, far from an ordinary introduction of motive. The court must be especially sensitive to the introduction of motive with racial connotations. When a court is dealing with narcotics use as motive for robbery, or a previous direct threat by a defendant against a victim as an indication of ill-will and motive, then there is a clear relationship between the defendant and the motive. There is no such relationship here.
The New Jersey Supreme Court also rejected petitioner's claims that the interjection of the racial revenge motive, and the state's summation of that motive, was an unacceptable appeal to racial prejudice, and as such violated their due process rights to a fair trial.Carter IV, at 107, 449 A.2d 1280. The summation read, in part:
PROSECUTOR: Strand Number Five: Motive. Defense counsel say that is a sensitive issue and it is. It's a very sensitive issue. None of us like to admit that things like race prejudice and anger and hate for people because of the different color of their skin exists in this world. We avoid it. We teach our children the contrary. We support civil rights. We support courses in our schools. We bear in mind the words of Reverend King, in which he had a dream of a day where people would judge his children by the quality of their character, not by the color of their skin. Now ladies and gentlemen, we don't live in that world yet and we certainly didn't live in that world in 1966. It was a world and is a world filled with people who *540 hate. You may remember in the voir dire a number of questions were asked of you about racial prejudice. We wanted a jury which is free from racial prejudice. But, we recognized in making those questions that not everybody is free from racial prejudice and, of course, we know that no group, no class is immune from hate and we know that revenge is one of the most powerful motives that any human being can have. We look around the globe and see it everywhere. We see Greeks and Turks and ...
COUNSEL FOR ARTIS: Your Honor, I object to Greeks and Turks and things outside this courtroom.
THE COURT: No. Counsel is making a point. He is illustrating a point by describing as other counsel did. I will permit that.
PROSECUTOR: We see hate and anger and revenge there and we see people in Ireland fighting because of religion and we know that in 1966 there were many blacks with legitimate grievances and some blacks and some whites did not act as law-abiding citizens. It's an area which everybody goes into with a great deal of trepidation, but does it mean we should hide, look away; we should say, well these motives are too repulsive, too ugly, we don't want to deal with them. We have to look at them. We have to analyze them. And I suggest to you, as much as you may want to look away, as much as you may want to say it couldn't have happened for that reason, it did happen for that reason. What other reason could it have happened for?
(45T198-200) (emphasis added)
D. Analysis of the State's Evidence
The spoken question begets the unspoken question: "What reason other than the defendants being black and the bartender white?" Even assuming, arguendo, that the government's version of events purporting to evince racial motive was supported by the evidence, the court concludes that the evidence submitted was insufficient to justify injection of such a contention. In essence, the prosecution was permitted to argue to the jury that defendants who were black were motivated to murder total strangers solely because they were white. Although racial revenge may indeed be a motive for murder, its highly inflammatory and prejudicial effect upon a jury requires that its existence be supported by sufficient and competent evidence. Furthermore, that evidence must be attributable and relevant to the defendants on trial and not based upon a stereotyped supposition as to how a particular group would react in a given situation. Therefore, in analyzing the facts in this case it is essential to distinguish the acts and knowledge of the defendants from those of other persons or groups of persons.
The prosecution, in support of its "racial revenge motive" relied upon what it has termed "twenty points" of evidence. (31T55-77) Of those points, only the following pertained directly to petitioners on the issue of motive:
1. Petitioners knew of the murder of Holloway, a black man.
2. Petitioners were friends of Edward Rawls, Holloway's stepson.
3. Carter extended his condolences to Rawls prior to the murders at the Lafayette Bar.
4. Carter testified before the grand jury that there was talk that night in the black community of a possible "shaking" (i.e. rock and bottle throwing).
*541 The balance of the purported evidence as to racial revenge might support the prosecutor's theory in general as to motive, but had little relationship to the petitioners. Indeed, it is difficult to fathom some of its admissibility as against them. For instance:
1. The three other Lafayette Bar patrons were shot only because they were witnesses to the killing of Oliver; there was no evidence of robbery. At best this evidence is consistent with the prosecutor's theory but adds nothing to implicate the petitioners.
2. There was a large and angry crowd gathered outside the Waltz Inn shortly after the shooting. It is difficult to understand how such a gathering could be imputed to petitioners. There is no evidence that they were present or even knew of such a gathering.
3. The Lafayette Bar had primarily white patrons and was an ideal target of the anger of the black community. Again there is no evidence that petitioners had any knowledge of the purported history of the Lafayette Bar or ever expressed any animosity towards it.
4. Rawls went to police headquarters and demanded action, became agitated and was ordered to leave. There is no testimony that petitioners knew or learned of this incident, and again it is difficult to comprehend how such activities could have been considered in assessing their motives.
If these and the other factors unrelated to petitioners are deleted from the government's equation, as this court concludes they should have been, what remains is simply the following: A white man killed a black man. Petitioners were friends of the stepson of the victim and expressed their condolences. They heard there might be a "shaking." Ergo they set out to murder four strangers solely because they were white as an act of revenge, and notwithstanding that one of the alleged murderers was well known in the community and easily recognizable.
Underlying the prosecutor's theory and summation is the insidious and repugnant argument that this heinous crime is to be understood and explained solely because the petitioners are black and the victims are white. Without that unacceptable assumption, the prosecution's theory of racial revenge becomes a thin thread (rather than the rope referred to in the prosecutor's summation) of largely irrelevant evidence and impermissible inferences.
The foregoing analysis is predicated upon a view of the prosecution's case in its most favorable light and accepting as true and proven each element of the prosecution's racial revenge theory. While this in itself would constitute sufficient grounds for granting the writ, the court is even further convinced of the correctness of its conclusion by the significant contradictory evidence which permeates the record on this issue. While this conflicting evidence as to motive is not in itself a constitutional infirmity, its presence further undermines the validity of its submission to the jury. For example, the motive in the Waltz Inn killing was a dispute over money, having nothing to do with race (33T120). While there was testimony at trial by two police officers that there was a crowd of 25 to 30 "unruly and angry" people outside the Waltz Inn, the atmosphere among the crowd was contradicted by another witness, Clarence Carr, who testified that while the group, which included whites, was upset, there were no racially derogatory terms used by the crowd, nor was anyone urging mob retribution on Conforti as he was brought through the crowd by police (41T18-19). Similarly, just as two police officers testified that Rawls appeared at police headquarters and promised vigilante action if the police did not bring his stepfather's killer to justice, another witness, William Johnson, said that Rawls made no threats or disturbance, but was simply making an inquiry. (34T31)
The testimony concerning the racial composition of the bars; that the Lafayette Bar was "white" and that Oliver, the bartender, refused to serve blacks, was vital to the question of motive. However, the police officer who testified that he had been *542 called to the bar on more than one occasion when Oliver refused to serve blacks also testified that he had seen blacks being served at the Lafayette Bar. (17T81) Another witness, Ronald Ruggerio, a white man who lived a few doors down from the bar, testified that he had seen blacks from the neighborhood patronizing the bar and had seen Oliver serving them. (40T130,-131)
The state asserts that the lynchpin of its racial revenge theory came from Carter himself, when he testified before the Grand Jury in 1966. Carter testified that there was talk among the black community of a "shaking", which he said meant to include rock and bottle throwing, but not murder (36T154). The state took Carter's definition further, implying that he meant murder, and stating in its summation that the only reason Carter mentioned the talk of "shaking" to the Grand Jury was that he knew the police were aware of it. (45T203) There was no evidence to support this, however.
Most significantly, there was no direct evidence ascribed to the petitioners to support the racial revenge motive. While Carter and Artis were shown to have known Rawls, and Carter conceded in his grand jury testimony meeting briefly with him at least one time that night to express his condolences, there was no evidence that either petitioner knew that it was a white man who killed Holloway, or, had they known, that they would have reacted in such a vicious and violent manner. There was evidence that the petitioners had cordial relationships with white people, socially and professionally, and that Carter lived in an integrated neighborhood and trained at an integrated gym. (39T135-136) Several white character witnesses testified on Artis' behalf. (43T113, 116, 118)
The petitioner's knowledge of a mob gathering or even threatened retaliation cannot be imputed to them unless they did more than know about it. Are the actions of a mob imputed to members of a particular race just because they witnessed or learned of the actions of such a mob? If defendants had learned of mobs, or threats, or talk of a "shaking" or retaliation from news reports rather than on the streets could such knowledge alone be used against them? Even assuming that petitioners had learned there was to be an indiscriminate killing of whites, how can such knowledge provide a basis for establishing their personal motives?
The final element of the state's theory was Carter's "search" for weapons. The prosecutor argued that Carter, swept up in the talk of a "shaking," went with his friends, including Neil Morrison, to the apartment of another friend, Annabelle Chandler, to confront her about a story she had told Carter about Morrison allegedly stealing guns from Carter's training camp a year earlier. (36T164) The state maintained that Carter somehow then found the guns, and after several meetings with Rawls, took part in the killings. The meeting at Mrs. Chandler's apartment was confirmed in the record, but Carter did not press the matter, because Mrs. Chandler was seriously ill. (36T147-148; 39T96) There is no testimony that Carter obtained the weapons at the apartment. But the links of this evidentiary chain are corroded. There was no evidence that Carter found the weapons. There was no evidence that Carter's conversation with Rawls went beyond a simple condolence, yet the prosecution inferred private meetings with Rawls, *543 during which, presumably, the killings were planned. Further, there was testimony that the decision to go to Mrs. Chandler's apartment was made before anyone had heard about the shooting, because of Carter's chance meeting with Morrison, who was the former manager of Carter's training camp and who had recently been released from jail. (36T136) After the meeting at Mrs. Chandler's apartment, Carter and his party returned to the Nite Spot, where they had been drinking before. (39T285-286)
Even accepting the facts relied upon by the state and all reasonable inferences therefrom, the court concludes the "search for weapons" scenerio lends no support to the state's racial revenge theory. The alleged search for guns is no more probative of the petitioner's motive than the murders themselves. Motive answers the question: "Why?" To argue that the search for guns is evidence of motive begs the question, particularly where it appears that the "search" may have occurred even before petitioners knew of the shooting of Leroy Holloway (36T140-145).
Certainly motives can be derived from actions, but to suggest that searching for guns, finding them and then using them, is evidence of motive is to elevate a bootstrap argument beyond all reason. Evidence that petitioners looked for weapons to carry out their racial revenge is probative of their guilt but it cannot also serve as evidence of their motive in seeking the guns. Indeed, if there was truly support for the prosecutor's theory, then the motive would have been formed when the alleged search for weapons began. The fact of the search adds nothing to the evidence of motive.
E. The State's "Assumptions"
The petitioners allege that the state's theories rest on three "unacceptable and insupportable" assumptions, one articulated and two unarticulated. (Petitioner's brief p. 102). The articulated assumption was that a "shaking" meant a violent, murderous response by members of the black community. As the prosecutor said in his summation:
A few hours before 8:00, I believe, a man at the Waltz Inn was killed. A white man came in and blew his head off with a shotgun and some four or five or six hours later two black men came into a bar and put a gaping hole with a shotgun into the white bartender. Coincidence? We like to think so, but the facts don't add up to a coincidence. They don't add up to a coincidence at all. A shotgun used, the same type of weapon; a bar on the border between a black area and a white area; a bartender who Officer Unger and this was his beat, remember testified that there had been complaints regarding the bartender refusing to serve black persons. What a natural target, and of course, there is Mr. Carter's testimony before the Grand Jury which was read to you. There was talk of a shake among certain persons in the black community. There was talk of a retaliation.
(45T201-202) (emphasis added)
The first unarticulated assumption, the petitioners argue, is that blacks in general, and Carter and Artis in particular, would have such hostile attitudes toward whites and such a predilection toward violence that they would be likely to respond by indiscriminately killing three whites whom they never knew to avenge the death of a black man they had never met. The second unarticulated assumption is that Rawls was a necessary co-conspirator, and the prosecutor's theory was based on a series of conjectures which required the jury, in effect, to try and convict Rawls, who was never indicted or brought into court. The inferential leaps made by the prosecutor are virtually impossible without the unstated appeal to the jury that it is perfectly reasonable to expect blacks to commit murder when one of their own is attacked. The fallacious premise of the argument becomes self evident if it is reversed and applied toward whites. Would a jury be permitted to conclude that a white defendant would have expressed such violent and *544 indiscriminate rage without any evidence of personal racial animosity?
The evidence did not support the imputation of the racial revenge motive to Carter and Artis. There was no proof that Carter and Artis were black militants with an inclination to kill whites, nor that they had even the slightest hostility toward whites, only that Carter had heard there was unrest and heard there was talk of a possible disturbance. In fact, the only blatantly racial statement placed before the trial court was Bello's testimony that while he was being interviewed by a prosecutor's detective in October 1966, that detective referred to blacks as "niggers" and "animals." (21T14)
Moreover, the prosecutor acknowledged the necessity to establish the showing of personal hostility in a letter to the trial court judge on November 26, 1976:
Wigmore (3rd ed. 1940 § 118) says that the showing of motive is a two-step evidential process. The first step is showing that a particular emotion is a circumstance showing the probability of appropriate ensuing action. Hence, a showing of the hostility of the defendant toward the race of the victim should be a circumstance which makes the desired inference (i.e. the defendant killed the victim) more probable.
Defendants' Joint Appendix on Appeal of Convictions, Volume 4, p. 22(a).
During oral argument, the state conceded that there was no direct evidence to support the motive other than Carter's "shaking" reference before the initial grand jury (Transcript of Oral argument July 26, 1985, p. 71). The state's answer to this lack of direct evidence is that "actions speak louder than words"; that if the jury were to find that Carter was the individual who committed the killings, they could then attribute the motive to him. Id., at 70. This argument is convoluted and contrary to the prosecutor's own arguments that the jury should use the existence of motive to find that defendants had committed the killings.
F. The Standard of Review
In reviewing the conduct of a state prosecutor in a petition for writ of habeas corpus, this court is limited to the narrow scope of due process violations. Donnelly v. DeChristoforo, 416 U.S. 637, 642, 94 S. Ct. 1868, 1871, 40 L. Ed. 2d 431 (1974). Only those state trial court errors that constitute a "failure to observe that fundamental fairness essential to the very concept of justice," violates due process. Lisenba v. California, 314 U.S. 219, 236, 62 S. Ct. 280, 290, 86 L. Ed. 166 (1941), reh'g denied, 315 U.S. 826, 62 S. Ct. 620, 86 L. Ed. 1222 (1942) ("In order to declare a denial of [due process] we must find that the absence of that fairness fatally infected the trial; the acts complained of must be of such quality as necessarily prevents a fair trial." Id.) In federal habeas corpus reviews of state proceedings, it is essential to distinguish between ordinary trial error and that sort of egregious misconduct which amounts to a denial of constitutional due process. United States ex rel. Perry v. Mulligan, 544 F.2d 674, 678 (3d Cir. 1976), cert. denied, 430 U.S. 972, 97 S. Ct. 1659, 52 L. Ed. 2d 365 (1977). Several circuits have said that a prosecutor's appeal to racial prejudice can rise to an infirmity of that magnitude. See: Miller v. North Carolina, 583 F.2d 701 (4th Cir.1978); Kelly v. Stone, 514 F.2d 18 (9th Cir.1975); United States ex. rel. Haynes v. McKendrick, 481 F.2d 152 (2d Cir.1973); and Soap v. Carter, 632 F.2d 872, 877 (10th Cir.1980) (Seymour, J. dissenting), cert. denied, 451 U.S. 939, 101 S. Ct. 2021, 68 L. Ed. 2d 327 (1981). "Appeals to racial prejudice are foul blows and the courts of this country reject them." Withers v. U.S., 602 F.2d 124; Ross v. United States, 180 F.2d 160 (6th Cir.1950), cert. denied 344 U.S. 832, 73 S. Ct. 46, 97 L. Ed. 648 (1952).
The essence of an improper appeal to the jurors is that it is directed to passion or prejudice rather than an understanding of the facts and the law. Perry v. Mulligan, at 680. Such appeals to passion or prejudice threaten the impartiality of the jury, a fundamental component of a fair trial. An effective appeal to prejudice undermines *545 the guarantee that a defendant shall have his case decided according to evidence in the record, rather than on the basis of potential juror bias. Moreover, prejudicial argument by the prosecutor unfairly stacks the deck in the government's favor. Soap v. Carter, 632 F.2d at 877 (Seymour, J. dissenting).
In Haynes, the Second Circuit reviewed a district court judge's issuance of a writ of habeas corpus on the grounds that the prosecution's use of racially prejudicial remarks in summation violated the petitioner's right to a fair trial. The remarks were more overt than in the instant case, but no less impressionable. The prosecutor in that case made repeated references to "colored people" as an entity separate and apart from the jury. Such "racial prejudice can violently affect a juror's impartiality and must be removed from the courtroom proceeding to the fullest extent possible." Id., at 157. Judge Oakes, in his opinion, traced the Supreme Court's fair trial cases, beginning with Moore v. Dempsey, 261 U.S. 86, 43 S. Ct. 265, 67 L. Ed. 543 (1923) where the Court had overturned convictions on Fourteenth Amendment grounds when racial prejudice "was a major factor in the fiber of the trial." He similarly traced the equal protection cases, beginning with Strauder v. West Virginia, 100 U.S. 303 [10 Otto 303, 25 L. Ed. 664] (1879) and noted there is a point where a criminal trial is affected by racial prejudice, either in the underlying procedure or the atmosphere surrounding the trial, when an overlap of the due process and equal protection clauses could occur. "If there is anything more antithetical to the purposes of the Fourteenth Amendment than the injection against a black man of racial prejudice ... we do not know what it is." Id., at 159. Judge Oakes concluded that there was a strong probability of prejudice in the prosecutor's remarks, and defined that "probability" as the correct test "when the evidence of guilt is not overwhelming." Id.
In Miller, three black men convicted of raping a white woman petitioned for habeas corpus, arguing, inter alia, that racially inflammatory remarks in the prosecutor's closing argument were so prejudicial as to make a fair trial impossible. Again, the remarks were overt: that the victim could not have consented to sexual relations because "the average white woman abhors anything of this type in nature that had to do with a black man," Miller, at 704. The court said that where evidence is relevant but also prejudicial, the law requires that it not be received into evidence until it has been demonstrated that its relevance and probative value outweighs its collateral prejudicial effect. Id., at 706. "A prejudicial argument by the prosecutor poses a serious threat to a fair trial. Not only does it undermine the jury's impartiality, but it also disregards the prosecutor's responsibility as a public officer." Berger v. United States, 295 U.S. 78, 85, 55 S. Ct. 629, 632, 79 L. Ed. 1314 (1935). See also: American Bar Association Project on Standards for Criminal Justice, The Prosecution Function, § 5-8(c) (1974) ("The prosecutor should not use arguments calculated to inflame the passions or prejudices of the jury.")
In McFarland v. Smith, 611 F.2d 414 (2d Cir.1979), a prosecutor asked the jury to credit a black police officer's testimony because she was testifying against another black person. "To raise the issue of race is to draw the jury's attention to a characteristic that the Constitution generally commands us to ignore. Even a reference that is not derogatory may carry impermissible connotations, or may trigger prejudiced responses in the listeners that the speaker might neither have predicted nor intended." Id., at 417. While the court added that not all race-conscious arguments are impermissible, it added a cautionary note:
But given the general requirement that the race of a criminal defendant must not be the basis for any adverse inference, any reference to it by a prosecutor must be justified by a compelling state interest.
The New Jersey Supreme Court and the respondents in this action dismiss this entire line of cases as not on point, because, *546 they maintain, the prosecutor's advocacy attributed no qualities to a generalized class of blacks. Carter IV, 91 N.J. at 108, 449 A.2d 1280. This court respectfully but vehemently disagrees. The evidence, so overwhelmingly circumstantial, requires, in many places, inference upon inference to support the state's theory. The only means by which the prosecutor's theory makes sense, in light of that evidence, is if the jury makes the prejudicial racial assumptions referred to above.
Moreover, in the context of the conflicting evidence and the extreme danger of prejudice, the remarks uttered in summation confirm the "unarticulated assumption" referred to by petitioners. The prosecutor, in analogizing the longstanding racial or ethnic wars abroad and the accompanying television images of senseless violence to the depth of racial antipathy that existed in Paterson in June, 1966, without basis in the record, imputed the "powerful motive of revenge" on the entire black community, and thus on the petitioners. This despite the absence of any evidence of either petitioner having such racial hatred. In sum, the prosecutor's theory invokes race for a purpose that has very slight or uncertain logical validity, and does so at a distinct risk of stirring racially prejudiced attitudes. McFarland v. Smith, 611 F.2d at 419.
The court must next consider whether injection of the racial motive into the case was harmless error under Chapman v. California, 386 U.S. 18, 87 S. Ct. 824, 17 L. Ed. 2d 705 (1967), reh'g denied, 386 U.S. 987, 87 S. Ct. 1283, 18 L. Ed. 2d 241 (1967). In Chapman, the Supreme Court stated that some constitutional errors are so unimportant and insignificant that they may, consistent with the Constitution, be deemed harmless, not requiring automatic reversal. "The question is whether there is a reasonable possibility that the evidence complained of might have contributed to the conviction," Id., at 23, 87 S. Ct. at 827 (emphasis added). It is certainly arguable that an appeal to racial prejudice of the kind which occurred here violates that category of constitutional rights so basic to a fair trial that their infraction can never be treated as harmless error. Id. Such cases include coerced confessions, denial of counsel or lack of an impartial judge. Id., at n. 8. In Miller v. North Carolina, the Fourth Circuit applied this standard of automatic reversal where the prosecutor made a blatantly racist comment. But this court need not reach the issue of whether that absolute standard has been met here; the probability of prejudice was sufficiently great and the case sufficiently close that it was reasonably possible, indeed probable, that the theory of a racial revenge motive and its submission to the jury contributed to the conviction.
G. Conclusion Racial Revenge Motive
To support this conclusion the court need look no further than the prosecutor's own argument to the trial court. In proffering the so-called "twenty points" of evidence to support the racial revenge motive, the prosecutor said:
if the state does not attempt to prove motive in this case, the state will be handicapped because these defendants will have every right to argue to the jury at the conclusion of the case that motiveless murders do not occur, that the state has not addressed any evidence of motive and for that reason these defendants did not do the crime for which they are charged.
Thus, the prosecution itself concedes that the racial revenge theory was essential to its procuring a conviction and without it, its case was in jeopardy. Certainly it should be estopped to now argue that it was not critical in the jury's determination of guilt.
Obviously, the death of the stepfather of the petitioners' friend, standing alone, would never explain why petitioners would shoot four innocent persons who were strangers to them. Notwithstanding the lack of any evidence that petitioners had a background of racial animosity against whites or had any such feelings after the specific death involved, the prosecutor was *547 permitted to render the illogical logical, by relying upon petitioners' blackness and the victims' whiteness. Thus the jury was permitted to draw inferences based solely on the race of the petitioners and the victims.
An appeal to racial prejudice and bias must be deplored in any jury trial and certainly where charges of murder are involved. For the state to contend that an accused has the motive to commit murder solely because of his membership in a racial group is an argument which should never be permitted to sway a jury or provide the basis of a conviction.
THE BRADY VIOLATION
A. The Harrelson Polygraph
As the 1976 retrial approached, the state faced the prospect of presenting its case without either of its crucial "eye-witnesses" from the 1967 trial. Arthur Dexter Bradley, who in 1967 had identified Carter as being present outside the Lafayette Grill, recanted his testimony in 1974. But more importantly, it appeared the state would not have Alfred Bello, the primary identification witness, who testified at the 1967 trial that he saw both Carter and Artis, with weapons, "on the street" outside of the tavern immediately after the killings.
Bello began retreating from his original testimony in a 1974 recantation and he continued revising his story until it became unrecognizable. At a recantation hearing in October, 1974, he said he could not identify either Carter or Artis as being near the Lafayette Grill after the shootings. Bello told recantation versions of his story to newsmen Raab and Levinson. In addition, during August 1975, Bello made a series of tape recordings with friends Melvin Ziem and Joseph Miller during which he told several additional versions of the events following the Lafayette Bar killings. On October 30, 1975, Bello said in an affidavit that he was "in the bar" at the time of the shootings, and that immediately afterward he ran outside, where he saw Carter and Artis. Bello said he did not remember seeing the petitioners in the bar and he did not believe they were the trigger men, but thought they were involved nevertheless. (20T175-178). In another affidavit the next day, Bello said that he did not see Carter and Artis with weapons or in the bar. (20T178). Bello told an Essex County grand jury seven weeks later that he saw two different black men in the bar, while he saw Carter and Artis outside. (Remand hearing 5/19/81 H79, 115-120). In June 1976, Bello told essentially the same story to the Passaic County prosecutor's office.
Understandably concerned, the state arranged for a polygraph examination by Professor Leonard H. Harrelson on August 7, 1976. The purpose of the examination, according to the then prosecutor, was to evaluate Bello's credibility, including "whether he was telling the truth or lying when he said he was in the bar at the time of the shooting," Carter IV, 91 N.J. at 133, 449 A.2d 1280. As a result of the polygraph, Harrelson concluded that Bello was telling the truth when he said he was in the bar shortly before and at the time of the shooting (5/19/81H79,131). He also concluded that Bello saw Carter and Artis outside of the bar after the shooting. Id. The same day, he gave an oral report of his findings to Passaic County Assistant Prosecutor Kayne and Chief of Detectives DeSimone. DeSimone told Harrelson that his conclusion was impossible; that Bello could not have been inside the bar at the time of the shooting. Id. Bello ultimately testified for several days at the 1976 trial, but the subject of the lie detector test was not brought out before the jury.
Harrelson testified at a post-trial remand hearing that he told the prosecutors his report was not tentative or preliminary and although he would not change his opinion, he would follow the verbal report up with a written "final" report. (5/19/H131-134) On August 11, 1976, Harrelson gave the same verbal report by telephone, and two weeks later he sent his written report to the prosecutor. The confusion that ensued was aptly described by Justice Clifford in his incisive dissent to the opinion of the New Jersey Supreme Court:
*548 For reasons that can charitably be described as unfortunate, in his later written report of his test the polygraphist summarized his findings with the opinion that Bello's 1967 trial testimony (which contained the "on-the-street" version) was "true": unfortunate, because Harrelson had never read Bello's 1967 testimony and no representative of the prosecution had enlightened him as to that testimony and hence he was plainlyand grievously mistaken as to the location from which Bello said, in 1967, that he had witnessed the events before and after the slayings; doubly unfortunate, because although the State continued to promote the notion that Harrelson's "in-the-bar" conclusion was only tentative [footnote omitted], Harrelson specifically and adamantly insisted that he never used those or any similar words or ever made the statement to "anyone at all on the face of the earth that [he] was unsure of Bello's test results * * *"; and, most unfortunate of all, because the prosecution never told the defense the critical finding of Harrelson's test that Bello was in the bar.
91 N.J. 86, at 134, 449 A.2d 1280.
B. The Brady Request
It is undisputed that before the 1976 trial, the defense made a specific request for all information concerning Bello's polygraph tests. Carter III, 85 N.J. at 313, 426 A.2d 501. The prosecution, believing the oral report to be preliminary, never told the defense about it before trial. Carter IV, 91 N.J. at 111, 449 A.2d 1280. The petitioners appealed this alleged violation of the Brady rule to the New Jersey Supreme Court, which remanded the case to the trial court for a determination of whether the violation had occurred, and to explore precisely what use the prosecution made of Harrelson's report in confronting Bello prior to the second trial. The trial court found that although the prosecution had "technically" failed to turn over information regarding Harrelson's oral report, that failure was justified under the circumstances. Carter V, at 110, 449 A.2d 1280. The New Jersey Supreme Court, on the other hand, unanimously agreed that a Brady violation had occurred. Id., at 112, 449 A.2d 1280. But by bare 4-3 vote majority, the court ruled that Carter and Artis had failed to overcome the standard of materiality required to show "the suppressed evidence might have affected the outcome of the trial." Id., (quoting United States v. Agurs, 427 U.S. 97, 104, 96 S. Ct. 2392, 2398, 49 L. Ed. 2d 342 (1976)). The majority ruled that the suppressed evidence was merely cumulative. Carter IV, 91 N.J. at 118, 449 A.2d 1280. The dissent, however, felt strongly that the evidence was material:
A more egregious Brady violation than the one presented by this case is difficult to imagine. One need not go so far as to impugn the motives of the prosecution in order to reach that conclusion, for it can just as easily be attributed to an appalling lack of basic communicative skills on the part of the principal polygraphist and various members of the prosecution team. But whether the circumstances originate in unworthy motives, colossal bungling, or plain dullness of comprehension, the fact remains that the misunderstandings thus created have proven to be costly indeed: the State witheld from the defendants material evidence favorable to them in connection with the Harrelson polygraph and, unknown to defendants and their counsel, compounded the error by using the mistaken and erroneous polygraph report to get the prime witness against the defendants to change his story again and go back to his original testimony given at the first trial. That all adds up to a deprivation of due process and requires a reversal of defendants' convictions.
91 N.J. at 133, 449 A.2d 1280.
C. The Applicable Standard
In Brady, the Supreme Court held that the suppression by the prosecution of evidence favorable to an accused (and requested by an accused) violates Due Process where the evidence is material either to guilt or punishment, irrespective of the *549 good faith or bad faith of the prosecution. 373 U.S. at 87, 83 S. Ct. at 1196. Accepting the conclusion of the New Jersey Supreme Court that Harrelson's oral report was withheld by the state, the court now turns to the issue of whether that non-disclosure is "material" under the Supreme Court's most recent Brady analysis, United States v. Bagley, ___ U.S. ___, 105 S. Ct. 3375, 87 L. Ed. 2d 481 (1985).
The Court first extended the Brady rule beyond exculpatory evidence to include impeachment material in Giglio v. United States, 405 U.S. 150, 92 S. Ct. 763, 31 L. Ed. 2d 104 (1972), in which the government failed to disclose evidence of a promise made by the government to its primary witness. Under the Giglio standard, a new trial is required if the testimony could, in "any reasonable likelihood," have affected the judgment of the jury. Id., 105 S. Ct. at 3388, (quoting Giglio, 405 U.S. at 154, 92 S.Ct. at 766). Subsequently, in United States v. Agurs, supra, the court distinguished three situations involving post-trial discovery of information favorable to the accused that had been known to the prosecution but unknown to the defense. The first situation involves the prosecution's knowing use of perjured testimony, where the verdict must be set aside if there is "any reasonable likelihood" that the false testimony would have affected the judgment of the jury. 427 U.S. at 112, 96 S. Ct. at 2401. The second situation involves the other extreme: where the defendant submits a "general request" or no request for information. In order to obtain a reversal in that circumstance, the Court held the defendant must show the omitted evidence "creates a reasonable doubt that did not otherwise exist." Id. The third situation, the middle standard, occurs when the defense makes a specific request for information and the prosecution fails to disclose it. Under this standard a reversal is warranted if the omitted evidence "might have affected the outcome of the trial." Id., at 104, 96 S. Ct. at 2398.
The issue in Bagley, as in the instant case, concerns the standard of materiality to be applied in determining whether a conviction should be reversed because the prosecution failed to disclose requested evidence that could have been used to impeach a government witness. ___ U.S. ___, 105 S. Ct. at 3375. Bagley, convicted of narcotics and firearms violations, alleged that the government withheld evidence that its two principal witnesses had signed during the investigation which stated the government would pay the witnesses for information they furnished. The suppressed documents could have been used by the defense to further impeach the witnesses. In Bagley, the Supreme Court adopted a single standard for "no request," "general request" and "specific request" cases of prosecutorial failure to disclose evidence favorable to the accused. The court relied on a rule first enunciated in Strickland v. Washington, 466 U.S. 668, 104 S. Ct. 2052, 80 L. Ed. 2d 674, reh'g denied, ___ U.S. ___, 104 S. Ct. 3562, 82 L. Ed. 2d 864 (1984):
The evidence is material only if there is a reasonable probabability that had the evidence been disclosed to the defendants, the result of the proceeding would have been different. A reasonable probability is a probability sufficient to undermine confidence in the outcome.
___ U.S. at ___, 105 S. Ct. at 3384
The "single standard" enunciated by the Court is somewhat semantical, however, because in order to assure that the standard has sufficient flexibility, certain factors must be weighed in determining *550 whether there is a "reasonable probability." The Court noted, for example, that the more specific the request for evidence "putting the prosecutor on notice of its value," the more reasonable it is for the defense to assume from the nondisclosure that the evidence does not exist, and to make pretrial and trial decisions on the basis of this assumption. Thus Bagley's instruction gives birth to yet another sliding scale: the more adverse the effect of the prosecution's failure, the more likely it is that the suppressed evidence will undermine confidence in the outcome.
[T]he reviewing court may consider directly any adverse affect that the prosecutor's failure to respond might have had on the preparation or presentation of the defendant's case. The reviewing court should assess the possibility that such might have occured in light of the totality of the circumstances and with an awareness of the difficulty of reconstructing in a post-trial proceeding the course that the defense and the trial would have taken had the defense not been misled by the prosecutor's incomplete response.
___ U.S. at ___, 105 S. Ct. at 3384
In light of its decision in Bagley, the Supreme Court vacated and remanded United States v. Oxman, 740 F.2d 1298 (3d Cir.1984), U.S. v. Pflaumer, ___ U.S. ___, 105 S. Ct. 3550, 87 L. Ed. 2d 673 (1985), which dealt with many of the same Brady issues. During the pendency of this habeas petition, the Third Circuit issued its revised opinion in that case, retitled United States v. Pflaumer, 774 F.2d 1224 (1985). Oxman/Pflaumer involved federal convictions for mail fraud and conspiracy where the government withheld specifically requested impeachment evidence that a government witness implicated in a check-kiting scheme had been given the promise of use immunity for his cooperation and testimony. The district court ruled that the testimony of the witness in question was "merely cumulative" of the other testimony at trial. Pflaumer, at 1229. After analyzing the nature and context of the testimony and applying the middle-level Agurs standard, the district court concluded that the nondisclosure "would not have affected the outcome of the trial." Id. In Oxman, a divided Third Circuit panel reversed the district court, holding that the information withheld was significant impeachment evidence.
In reconsidering the case in light of Bagley, a still divided Third Circuit panel reversed itself, holding that the district court "correctly focused on the reasonable probability that, had the [agreement with the witness] been disclosed, the result would have been different. Id. The court said, "the government's failure to disclose the immunity agreement in the context of the facts and full record in this case does not undermine our confidence in the verdict." Id., at 1230. The court looked at the evidence of Pflaumer's guilt, both direct and circumstantial, from multiple and corroborating sources, and downplayed the importance of the witness in comparison to the incriminating testimony from other witnesses in the case. Id., at 1230. The court also was careful to distinguish situations, such as the instant case, where witnesses play a more pivotal role in the evidence. For example, the court distinguished United States v. McCrane, 547 F.2d 204 (3d Cir.1976) (per curiam) which involved suppressed letters written by prosecutors on behalf of their sole witness and Bagley, which involved one of two incriminating government witnesses.
In the case before the court, the suppressed oral report goes to the veracity of Bello, the only state witness who could identify the petitioners at the scene of the crime. It is crucial; without that cracked and shaky pillar to support it, the balance of the government's case would probably come crashing down. And although Bello withstood several days of cross-examination, it is unlikely that his credibility would not have been totally destroyed by this disclosure. Obviously then, the greater the effect the undisclosed report can be shown to have had on Bello's credibility, the more material the concealment would become.
*551 D. The New Jersey Supreme Court Findings
The New Jersey Supreme Court dissent approached the effect of the non-disclosure in two ways: First, what use the prosecutor made of the ostensible, but unintended conclusion in Harrelson's written report (confirming the 1967 testimony) that Bello had been "on the street" during the shootings, and second, what use the defendants would have made of the intended, but undisclosed oral report. Carter IV, 91 N.J. at 135, 449 A.2d 1280. Before examining the first question, the court must determine what weight to give to factual findings by the court below in a habeas corpus proceeding.
Under 28 U.S.C. § 2254(d), a determination after a hearing on the merits of a factual issue made by a state court of competent jurisdiction [and] evidenced by a written finding, written opinion, or other reliable and adequate writing indicia, shall be presumed to be correct unless the applicant for a federal writ of habeas corpus can establish one of the causes for exception listed in the statute. The § 2254(d) presumption of correctness does not apply to legal questions or mixed questions of law and fact, thus opening the holding to review on collateral attack. Cuyler v. Sullivan, 446 U.S. 335, 342, 100 S. Ct. 1708, 1714, 64 L. Ed. 2d 333 (1980). On remand from the New Jersey Supreme Court, the trial court made numerous findings of fact, among them:
(13) The prosecution did not cause Bello to recant his recantation through coercion, deceit, or threats of prosecution for perjury; nor did the state misrepresent to Bello the results of the polygraph test to get Bello to back to the "on the street" version;
(14) Bello recanted the recantation when he told Harrelson during the polygraph pre-test interview that Raab, Hogan and Levinson pressured him into recanting the 1967 trial testimony;
91 N.J. 86, 119, 449 A.2d 1280 (quoting trial court's Opinion on Remand p. 54).
The Supreme Court majority upheld these findings, which go to the first part of the court's inquiry: what use the prosecutor made of Harrelson's oral report. The trial court's determination as to what use the prosecutor made of the written report in convincing Bello to return to his original story is a finding of fact that assumes a presumption of correctness. Cuyler v. Sullivan, supra; Swain v. Pressley, 430 U.S. 372, 97 S. Ct. 1224, 51 L. Ed. 2d 411 (1977).
Petitioners, on the other hand, argue that the presumption of correctness does not apply because they did not receive a "full, fair, and adequate hearing," under § 2254(d) (6). They continue to allege that the withholding of the report was part of the manipulation of Bello by the prosecution. The Supreme Court dissent concluded that the prosecution knew that Harrelson's written report contradicted its findings regarding the vantage point from which Bello supposedly made his pivotal observations. The dissent concluded that prosecutors "concealed the fact that the "wrong" test result was fed to Bello to break him." Carter IV, 91 N.J. at 137, 449 A.2d 1280. This conclusion does not conflict with those made by the trial court, the dissent noted, because after Bello recited pressures to recant, he thereupon told Harrelson, directly contrary to his 1967 testimony, that he was "in the bar" when the shooting took place. Id. The dissent also concludes, based on the testimony of an assistant prosecutor at the remand hearing and a letter from the prosecutor to the state attorney-general on November 22, 1976, that sometime after he had been subjected to the polygraph technique, Bello was confronted with the unintended "on-the-street" result of Harrelson's report, and that this confrontation caused him to return to his 1967 "on-the-street" version. Id., at 136, 449 A.2d 1280. Thus, notwithstanding the factual findings of the New Jersey courts, defense attorneys could still have focused on the prosecution's use of the suppressed information to continue to hammer away at Bello's malleability. They could have also chipped away at a main *552 theme of the prosecution that Bello was manipulated away from his original testimony because of some sort of bribes from people in the defense camp.
E. The Scope of Review
The court has a much wider purview in considering whether the standard of materiality enunciated by the New Jersey Supreme Court was correct. In Chaney v. Brown, 730 F.2d 1334 (10th Cir.1984) cert. denied, ___ U.S. ___, 105 S. Ct. 601, 83 L. Ed. 2d 710 (1984), the defendant, convicted of murder and sentenced to death by a state court, sought a writ of habeas corpus, claiming that the prosecution suppressed Brady evidence. The Oklahoma Court of Criminal Appeals concluded that the withheld evidence was not material as to guilt or punishment. Id., at 1341. While the Tenth Circuit also grappled with the state court's use of an incorrect standard for determining materiality, it held that the determination of whether withholding of the evidence affected the verdicts so as to infringe due process rights is not a factual finding to which § 2254(d) applies. Id. at 1345-46. The court cited Davis v. Heyd, 479 F.2d 446, 451 (5th Cir.1973):
The question of materiality present in cases in which the accused complains of prosecutorial suppression of material evidence is ... [a] mixed question  of law and fact calling ultimately for a legal determination ... and therefore [is] not entitled to a § 2254(d) presumption of validity.
Id. at 1346.
Respondents urge that the ultimate finding of the trial court, that the non-disclosure of the oral report "in no way would or could have affected the outcome of the second Carter-Artis trial" (Opinion on Remand p. 74) should be entitled to great weight. The finding is clearly labelled as a conclusion of law in the opinion on remand and was subsequently rejected by the New Jersey Supreme Court as it is by this court. Therefore, the court turns to what use the petitioners could have made of their knowledge of the Harrelson oral report at trial.
The New Jersey Supreme Court majority determined that the suppressed evidence was "merely cumulative." Carter IV, 91 N.J. at 118, 449 A.2d 1280. It quoted the trial court's determination that the jury was fully aware of Bello's criminal background and unlawful activities; that he was portrayed before the jury as a perjurer, liar and thief. "Given the length and scope of the defense's attack on Bello's testimony, there is no reasonable likelihood that further impeaching evidence would have affected the outcome of the trial." Id. The court cited three cases where the suppressed evidence was merely cumulative. All are distinguishable from the instant case. In United States v. Goldberg, 582 F.2d 483 (9th Cir.1978), cert. denied, 440 U.S. 973, 99 S. Ct. 1538, 59 L. Ed. 2d 790 (1979) the government's case consisted primarily of the testimony of a fellow employee. The prosecutor had failed to respond to a specific request for production of the employee's notes, which could have been used to better illustrate his cooperation with the government in hopes of getting better treatment. But the notes merely amplified the agreement, which was already brought out at trial. There was nothing in the suppressed material which was inconsistent.
In DeMartino v. Weidenburner, 616 F.2d 708 (3d Cir.1980), the defendant had been found guilty of accepting a bribe, and the prosecutor failed to make available allegedly exculpatory material by a third party to a government investigator. However, the defendant already had presented other witnesses who could provide the same exculpatory information. In addition, the court noted that the statements were made after the defendant learned that he was the target of an investigation, reducing its evidentiary value. Id., at 711. Finally, Zeigler v. Callahan, 659 F.2d 254 (1st Cir.1981) involved suppressed information about the criminal history of the key government witness. The court held, unlike the instant case, that the information contained in the undisclosed statements did not relate to the *553 substance of the witnesses' testimony. Id., at 266-67.
F. The Materiality Standard
Petitioners claim that they could have made effective use of the lie detector test had they known the true state of affairs. (6/3/81H29-32). The defense could have had additional evidence to undermine the credibility of Bello as an eyewitness during Wade/Stovall hearings before the trial court, challenging his ability to identify the defendants and describe their movements if he was in the bar rather than outside the bar. Furthermore, the defense could have attacked the credibility of DeSimone and Kayne for concealing the polygraphist's conclusion that Bello was in the bar. Finally, the defense attorneys could have focused more heavily on Bello's manipulation and on his total unbelievability in their summation.
Use of this information was far more than another attack upon Bello's credibility. If counsel for petitioners had the relevant information, they would have had the means to convince the jury that Bello selected one of several versions, possibly all untrue, merely because he mistakenly believed it had been confirmed by a polygraph test. If he had been bribed to select one of the versions, certainly it could not be argued that such evidence would have failed to have had a profound impact upon the jury.
What distinguishes this situation from those in which attacks upon credibility are deemed merely cumulative is the fact that Bello was permitted to urge his "final" version as being motivated by the truth rather than by circumstances having no relation to the truth whatsoever. The significance of this failure is not that it denied the defendants just one more opportunity to challenge the credibility of Bello. The witness had been confronted with his several versions and thus his credibility was placed squarely in issue. But the petitioners were deprived of the opportunity to demonstrate why the witness settled on this particular version and to bring out the role of the prosecution in that choice.
Absent the potent ammunition this knowledge would have afforded the defense, Bello was free to say that he had selected this one of several versions because it was "true." However, if defendants had known all of the relevant facts, they could have argued not only that the choice was predicated upon facts submitted by the prosecution, but that those facts were false. The jury could well have concluded then that if Bello had been told that the "in-the-bar" version or some other version was found true by the polygrapher, he would have told that version. From those circumstances the jury could have concluded that Bello's choice of the final version was not a decision based on truth but rather one influenced by a premise, and a false one at that, furnished by the prosecution.
Although a finding of intentional misconduct on the part of the prosecution enhances petitioners' argument, it makes no real difference under the factual circumstances of this case. Whether the conduct was deliberate or negligent, the consequences to the petitioners were the same: they were deprived of a vital opportunity to totally discredit the key and only eyewitness to the crime. Indeed, if the trial court knew and was satisfied that Bello finally selected one of his many versions merely because he was told that it was independently confirmed by the polygrapher (albeit mistakenly), it might well have stricken his entire testimony.
At trial, counsel for the petitioners feared opening the door to introduction of the lie detector results, believing it would show simply that Bello's 1967 version was truthful. Had they known about the conflicting oral report, and Harrelson's insistence as to its accuracy, their fears would have dissipated. The New Jersey Supreme Court majority downplays this stratagem theorizing that it would have involved introduction of the lie detector results, which it cited as potentially even more damaging than any possible added impeachment. Carter IV, 91 N.J. at 116, 449 A.2d 1280. *554 Respondents refer to the possibility as "insurmountable rehabilitation of the witness." (RB p. 98)
The petitioners argue that the state Supreme Court misstated facts in relation to the lie detector tests which are significant. For example, the court said that if laid out before a jury, the Harrelson test would have established the polygraphist "entertained no doubt at all that Bello was truthful when he identified the defendants as the murderers. Id. In fact, Bello never identified Carter and Artis as the murderers. (Appendix in Support of Carter's motion seeking order compelling withdrawal of Passaic County Prosecutor 37a-57a) Secondly, the court said Bello told Harrelson he saw Carter and Artis "outside the bar with weapons immediately after the shooting." Carter IV, at 120, 449 A.2d 1280. In the actual test, Bello was not asked, nor did he say that the two men had weapons, although he told Harrelson he saw them outside the bar. Appendix in Support, supra.
The dissent points out that if the state tried to show the truth of the polygraph, the defense could have shown its contradictions; the entire notion of polygraph evidence having validity in regard to Bello could have been laid to rest if the state tried to introduce yet another polygraph report it had obtained based on Harrelson's records which said Bello was truthful when he said he was "on-the-street" at the time of the shooting, while Harrelson maintained throughout that Bello's "in the bar" version remained correct.
Had the trial court been aware of all the facts, it could have formulated some way in which the polygraph reports could have been limited for the purposes of credibility without allowing the prosecution to use them to bolster their case. Numerous instructions or stipulations were available to avoid the polygraph results ever being presented to the jury, i.e.. Mr. Bello was told that a polygraph had confirmed his "on the street" version. Either: This was not true (or) This was a mistake.
Thus, to view the withholding of the oral report as merely cumulative "is to gloss over the essential nature of that inconsistency and misgauge its potential impact," Id., at 139, 449 A.2d 1280 (Clifford, J., dissenting). It is sufficient to undermine confidence in the outcome? If the outcome depends in large measure on Bello's testimony, how could it not? "Never before could defendants argue so persuasively that Bello was in all respects a complete, unvarnished liar, utterly incapable of speaking the truth." Id.
G. The Totality of the Circumstances
The court must now assess the adverse impact of the prosecution's failure to turn over the oral report in light of the totality of the circumstances. In its summation, the state divided its case against the petitioners into six "strands" of evidence. (45T147)
1. The identification of the killer's automobile by Patricia Valentine.
2. The identification of petitioners by Alfred Bello.
3. Circumstances surrounding the apprehension of the petitioners.
5. Location of a bullet and shotgun shell in the trunk of Carter's car.
6. False alibis presented by petitioners at the 1967 trial.
The court has ruled upon the use of the racial revenge motive and the reliability of Bello's testimony. Each one of the remaining strands is frayed.
1. Identification of the killer's automobile: Patricia Valentine, the witness who lived above the Lafayette Bar, testified at trial that the white car leaving the murder scene had taillights identical to those of the white 1966 Dodge Polara brought to the scene later by police with Carter and Artis inside. (15T145-77). She and officers at the scene testified that after she saw the taillights of Carter's car, she became hysterical and ran away. Mrs. Valentine also testified that the car had out-of-state license plates with yellow or gold numbers *555 and a dark blue background. Bello testified that he gave police a description of the car as a white, highly, polished vehicle, with triangular taillights, and further identified Carter's car at the scene as the "identical car" (19T182,195).
Petitioners argue that there is nothing to indicate from police reports that Mrs. Valentine identified the Carter car when it was brought to the scene (30T63). She testified that she saw only the rear of the car. (16T97) Moreover, her testimony that the taillights were identical was new to the second trial. In the first trial she said merely that the rear of the car was "similar." (16T100) Petitioners claim her description of the taillight configuration was identical that of a different Dodge model, the Monaco, which she herself identified as the model of the perpetrator's car in her 1966 grand jury testimony. (16T141-142) A Paterson police officer testified that she had identified the taillights of Carter's car at the scene, but conceded on cross-examination that there was no mention of this identification in his reports or at the 1967 trial. (30T63-67).
While Bello also claimed at trial to have identified the getaway car to police when they arrived at the scene, the police radio merely described the car as white with two black males inside (30T88). The petitioners also present conflicting evidence as to what information was originally given to police and broadcast. This has some significance because police had chased and/or stopped several white vehicles after the shooting. (40T97, 99, 101, 163).
2. Eyewitness Identification: At the scene Bello gave a statement identifying Carter's automobile but not the petitioners. (30T177-178, 34T175) Later he testified that he feared providing the information because he was on parole and had been a lookout in an ongoing burglary when the killings occurred. (19T196-97) After the shootings, he took money out of the cash register and gave it to Bradley, his accomplice in the burglary. Id.
At trial, Bello said he recognized Carter and Artis when they were brought to the crime scene (21T92-93), and that he would recognize Carter because he had seen him twice before, once in a boxing match while Bello was an inmate at a state reformatory, and another time at a Paterson bar (20T232-234, 240-241). Petitioners argue that Bello had never seen Carter at close range, and his descriptions began taking shape only after he saw the petitioners at the scene. (PB at 41, PRB at 19).
Bello's identification must be taken in context with the subsequent changes in his story (See: The Brady Violation, infra). His malleability was first demonstrated during an interview with two members of the prosecutor's staff. (PB at 43-47, citing Defendants' appendix on appeal 3DaT42, 43); and his motivation was crystalized when, shortly after the trial, police officers and others tried to help him obtain a $10,000 reward offered to persons providing information that led to the arrest and conviction of the Lafayette Bar killers (24T32, 34).
Bello continued to rely upon police to intercede on his behalf with the courts (29T59-64) until he was finally told, in 1974, that nothing more would be done for him. (29T67) It was at about that point that Bello began to talk to people in the Carter camp concerning a recantation. (37T23-24)
Bar patron Marins testified that the killer who wielded the shotgun had a mustache, but the respondents stress that Marins was in shock, and probably drunk. (15T152) Petitioners respond that Marin's testimony was extensively detailed and was relied upon by the state to determine the entry and exit of the killers.
Petitioners point out that Marins could not identify them the hospital (17T185). Later, he described both men as being about six foot tall, one light colored with a pencil line mustache. Neither was described as having a goatee. (17T178-222) Bar patron Tanis, at the hospital, gave varying descriptions, and the investigator who interviewed her said she told the same story as Marins: "Both men were Negros, and the one with the shotgun was about six *556 feet, slim build, light complexion and a pencil line mustache." (17T173). Petitioner Carter, at the time, was described at five foot seven inches, 160 pounds, very dark complexion, well-built, with a prominent goatee. Artis was described as being six foot one inch, with a slight to medium build, a medium complexion, and clean-shaven. (PB at 22). Before Bello came forward, the Passaic County detective who led the investigation told the first grand jury on June 29, 1966 that "the physical description of the two holdup men is not even close." (PB at 22, citing transcript Defendant's appendix on appeal 4DaT57)
Mrs. Valentine said the men wore sports jackets, while the one who entered the passenger side of the getaway car was wearing a fedora. (16T27-28). According to Bello's initial description, the first man was a black male wearing a fedora and sports jacket, thin, and about five foot eleven inches tall; the second man was thin and about the same height. (18T55-56; 30T24). When brought to the scene, Carter wore a white jacket with thin stripes, a red vest and dark pants and straw hat (30T92). Bello described the other man as wearing dark clothing (19T164), yet Artis wore a light blue sweater with his initials imprinted, a blue v-neck shirt and light blue pants (29T172; 42T217).
3. Movements of the Petitioners: About four minutes after the reported time of the shooting, Paterson police were chasing a white car speeding out of Paterson (30T86-87). Some ten minutes after the shootings, the same two policemen, having failed to stop the other white car, stopped the car carrying Carter, Artis and John Royster. They checked the car's registration, recognized Carter, and allowed the car to go on. (30T91-95). This was about 15-16 blocks from the murder scene. (30T120-121). Shortly after, a police description of the car was broadcast, the two officers again stopped the Carter car, this time even closer to the crime scene, this time with only the petitioners inside.
The respondents theorize that the speeding car and Carter's car were one and the same, and after it had escaped the police chase, the car stopped at the home of Edward Rawls, where Carter and Artis dropped off the murder weapons and changed some of their clothing. They then stopped at the Nite Spot to pick up Royster for alibi purposes. (RB at 32)
Earlier, the state charged that Carter had engaged in a search for weapons (See: Racial Revenge Motive, infra). The testimony corroborating Carter's latest whereabouts before the killings was relied upon by the state was that of Elwood Tuck, the manager of the Nite Spot, who had last seen Carter in his back room about 2:15 a.m.
Petitioners point out that the policemen who stopped his car could not identify it as the one they chased out of town (30T125-126). Another police officer testified that he saw the Carter car and asked both headquarters and the officer at the crime scene whether they wanted the car stopped, and he was told no both times. (31T166-168). The 1967 trial testimony of the officer in charge of the crime scene made no mention of a detailed description of the getaway car or an identification of Carter's car by Bello.
4. The Bullet and Shell: The state produced a live shotgun shell and a live revolver bullet, each respectively matching the caliber of the weapons used at the Lafayette Bar. (36T46-64) A Paterson police detective testified he found the shell, a .12 gauge "Western," in the trunk of Carter's car under boxing equipment and found a .32 caliber S & W long bullet on the floor of the front seat (35T127-130). A newspaper reporter friendly with the detective and present in the early morning hours at the police garage where the discovery took place testified that he saw the detective find the ammunition. (35T277-315) Two other people testified that the detective told them about the discovery soon afterward.
Petitioners argue that the detective who found the ammunition had not vouchered it with the police property clerk for five days. The significance of the delay, according to petitioners' theory, is that the detective either intentionally or unintentionally produced *557 ammunition found earlier that evening in the course of investigating the Holloway killing. The bullet found in the Carter car was brass cased, rather than copper coated like those found at the Lafayette Bar. (36T63, 75) Similarly, the shotgun shell found in the Carter car was an older model, with a different wad and color. (36T89, 90, 234, 236).
Moreover, in his reports on the Holloway murder, the detective had identified two type of shells found as Sears Sportsload and Western (35T144-160). The detective later amended his report to say there were no Westerns found after the Holloway murder. (35T144, 152)
Petitioners also dispute the accuracy of the interrogation of Carter in which he purportedly denied lending his car or knowing about the ammunition; two points upon which the state relies heavily. Petitioners note the statement was never seen or acknowledged by them (32T89, 140) and that the detective who interrogated them conceded destroying his original notes after reducing them to typewritten form. (32T96-97). The notes do not include any reference to Carter's whereabouts between 2 a.m. and 3 a.m., a topic one would expect to be the primary reason for the interrogation in the first place. The New Jersey Supreme Court criticized the admissibility of the notes, but concluded that the affirmative probative value of these oral statements was virtually nil. Carter I, 54 N.J. at 442-446, 255 A.2d 746.
5. The False Alibi: On the surface, the most damaging evidence against Carter may have been testimony at the 1976 trial concerning a false alibi fabricated by or on behalf of Carter at the 1967 trial. The alibi was meant to cover the time period shortly before and after the murders; from 2:15 a.m., when Tuck last saw Carter at the Nite Spot, through 2:40 a.m., when Carter was stopped in his car 15 or 16 blocks from the crime scene.
Welton Deary had testified in 1967 that he saw Carter at the Nite Spot with Catherine McGuire, her mother, Anna Mapes Brown, and William Hardney. He later testified that he was lying (27T116-119), and that he had actually seen Carter about 1 a.m. at another bar, Richie's Hideaway with Edward Rawls (27T147). He alleged that Carter's defense team had asked him to change his story. (27T 119-120)
Anna Brown testified in 1967 that Carter had driven the two women home from the Nite Spot about the time the killings took place. In 1976, she said that testimony was false and that Carter had driven the two women home another night. Catherine McGuire told a similar story, and also testified that she received a letter from Carter while he was in jail asking her to make sure her testimony covered him at 2:30 a.m. That letter and others, ruled inadmissible at trial because they were illegally obtained, were nevertheless permitted to be related to the jury in order to refresh the recollections of witnesses. (28a.m.T61-71).
William Hardney was not an alibi witness in 1967 because he was living in Washington, D.C. and was afraid to come to New Jersey, where he was a fugitive from non-support charges (27T51). Hardney testified in 1976 that Carter wanted him to back up the McGuire-Brown alibi by saying he saw Carter take the two women home from the Nite Spot about 2:15 a.m. (27T54)
Respondents further allege that Rawls and Neal Morrison, were part of a team of defense helpers taking witnesses to a motel that served as defense headquarters during the 1967 trial. There, the state says, defense counsel helped induce the false testimony (28a.m.117-133; 27T120; 27T185)
Petitioner Carter, however, stands by his original alibis, and points to changes in the lives of the witnesses that paralleled significant changes in their testimony from 1967 to 1976. (PB at 65). Deary, for instance, who was unemployed the night of the killings, subsequently became a Paterson Housing Authority policeman. (27T143). He was interviewed by prosecutors while on duty and in uniform. Another Carter associate, Merrit Wimberly, testified that Deary told him he had changed his testimony to protect his job (39T51) Wimberly, a *558 1967 alibi witness, repeated his 1967 testimony at the second trial that he saw Carter leave the Nite Spot at 2:15 a.m. with McGuire and her mother and return about five minutes later, and was still at that bar when Wimberly left at 2:30 a.m. (39T47). Catherine McGuire was engaged to a sergeant in the Paterson police department in 1976 (39T51). Prosecutors threatened her with perjury for her conflicting statements (28a.m.T113).
William Hardney testified that several Paterson detectives paid him a late night visit in Washington in October 1976, and questioned him for hours without allowing him to contact his lawyer. (27T75-79). He testified that the investigators returned to his home several times in the presence of his family, threatening to arrest him and his wife for obstruction of justice. Ultimately he was arrested and brought to Paterson during the 1976 trial. He was released only after he testified on November 27. (27T110). However, Hardney's presence at the Nite Spot the night of the killings was confirmed by Tuck, who was relied upon by the state as well as the petitioners (39T224).
Raymond Brown, the lawyer who represented Carter in 1967, testified that McGuire and his mother both made voluntary statements and were not pressured in any way. (42T22-33). Brown also disputed the accusation that he tried to change Deary's testimony as well. Artis' counsel at the 1967 trial, Arnold M. Stein (now a state Superior Court judge), confirmed Brown's testimony and the voluntariness of the women's statements. (42T130-133)
Finally, petitioners point to United States v. Burse, 531 F.2d 1151 (2d Cir.1976) to underscore the point that the jury should not have utilized disbelief of an alibi defense to lessen the state's burden of proving guilt by a reasonable doubt.
Petitioners emphasize the doubt surrounding respondents' theory of events by summarizing the respondents' best evidence and assuming, arguendo, that: Valentine identified the Carter car at the scene and at the police garage, that Bello identified Carter and Artis at the scene as well as identifying the car at the scene and at the police garage, that the police indeed found the ammunition in the car at headquarters and that the petitioners failed their initial lie detector tests and gave conflicting stories to the police. If this is all true, why were petitioners released without being charged? Why were they exonerated by the grand jury, and why did a Passaic County detective tell the first grand jury that the policeman who administered the initial lie detector test believed Carter was not a participant in the crime, although he had knowledge as to who was responsible? (Grand Jury transcript, reproduced PB, appendix, at B3)
Even at its strongest links, the government's chain of evidence has been substantially called into question by petitioners. Therefore, considering the totality of the circumstances, the court concludes that had the evidence withheld by the state involving Harrelson's oral report been disclosed to the defense, there is a reasonable probability that the result of the trial would have been different. Bagley, ___ U.S. at ___, 105 S. Ct. at 3384.
THE REMAINING CLAIMS
The court finds that the petitioners have shown, by a preponderance of the evidence, that (1) the state violated the requirements of the Brady rule by failing to disclose the results of a lie detector test given by the state to its only "eyewitness," and (2) that the state violated the Due Process rights of petitioners by improperly appealing to racial prejudice during the trial by arguing that the killings were motivated by racial revenge.
While the court has reviewed all the claims in the petitions, the two grounds cited above are, in the opinion of the court, the petitioners' most powerful arguments for the granting of the writ and are clearly supported by the record. Accordingly, the court finds it unnecessary to reach the merits of the remaining claims, Grounds 4, 6, 7, 8 and 10.
*559 EXHAUSTION OF STATE REMEDIES
In Rose v. Lundy, 455 U.S. 509, 102 S. Ct. 1198, 71 L. Ed. 2d 379 (1982), the Supreme Court ruled that 28 U.S.C. § 2254(b), (c) requires a federal district court to dismiss a writ of habeas corpus containing any claims that have not been exhausted in the state courts. This doctrine does not bar relief where the state remedies are inadequate or fail to "afford a full and fair adjudication of the federal contentions raised." Id., at 516, n. 7, 102 S. Ct. at 1202, n. 7, (citing Ex Parte Hawk, 321 U.S. 114, 64 S. Ct. 448, 88 L. Ed. 572 (1944)). The court noted that if the petition contained both exhausted and unexhausted claims, a prisoner could always amend his petition to delete the unexhausted claims rather than return to state court to exhaust the remainder of his claims. Id., at 520, 102 S. Ct. at 1204.
Carter amended his petition to delete the unexhausted claim in Ground 12. However, Artis argues that the delay by the New Jersey courts in adjudicating the claim in Ground 12 has been so extraordinary that it inflicts a manifest injustice upon him. Counsel for Artis has asked the court to declare that, because of this delay, the petitioner has, in effect, exhausted his state remedies. The state opposes this motion.
On November 1, 1985, during the pendency of this petition, the New Jersey Supreme Court denied the petition for certification by Carter and Artis. Therefore, the court finds that petitioner Artis has exhausted his state remedies, and is entitled to full consideration of his petition.
THE APPROPRIATE RELIEF
Under 28 U.S.C. § 2243, the court is instructed to dispose of federal habeas corpus proceedings "as law and justice require." "The prevailing practice under this mandate is to delay the petitioner's discharge so as to allow the state reasonable time in which to retry the petitioner if it wishes to do so." Ridge v. Turner, 444 F.2d 3, 5 (10th Cir.1971). Whitely v. Warden, Wyoming State Penitentiary, 401 U.S. 560, 569, 91 S. Ct. 1031, 1037, 28 L. Ed. 2d 306 (1971).
Immediate release from custody with prejudice is rarely awarded, and when it is, the appellate courts may find abuse of discretion. Yackle, Postconviction Remedies, § 141 at 530. Bromley v. Crisp, 561 F.2d 1351 (10th Cir.1977), cert. denied, 435 U.S. 908, 98 S. Ct. 1458, 55 L. Ed. 2d 499 (1978) (reversing a district court's order vacating a state judgment with prejudice). Hammontree v. Phelps, 605 F.2d 1371 (5th Cir. 1979), Ridge v. Turner, supra.
The killings that led to the petitioners' indictment and conviction occurred nearly 20 years ago, and to retry such conflicting events, further aggravated by dim memories, does not appear to serve the interests of justice. Moreover, to again use Bello as such a key witness, after his unbelievable series of recantations and recantations of his recantations, his complete and utter malleability at the hands of all parties, and his own sordid criminal history, would probably place his competency as a witness beyond the outer limits of due process.
Yet, even though a new trial may very well be a practical impossibility, this is a decision that in the interests of comity should be made by the state. In Gardner v. Pitchess, 731 F.2d 637 (9th Cir.1984), the district court granted a writ on the basis of ineffective assistance of counsel. In that case the magistrate recommended dismissal on the grounds that a fair retrial would be difficult. The Ninth Circuit ruled that the unfairness of a retrial was not properly before the district court, and could be better reviewed by the state courts if a retrial occurs. Id., at 640.
Therefore, the court will grant the writ as to both petitioners, mindful that the state can seek a retrial, but hopeful that constitutional considerations, as well as justice and compassion, will prevail. See: Clausen v. Clerk of Circuit Court, 537 F. Supp. 1233, 1237 (E.D.Wisc.1982), White v. Estelle, 554 F. Supp. 851, 859 (S.D.Tex. 1982).
The court does not arrive at its conclusion lightly, recognizing that it is in conflict with a decision of the highly learned and respected New Jersey Supreme Court (albeit a 4-3 decision). However, this court is convinced that a conviction which rests upon racial stereotypes, fears and prejudices violates rights too fundamental to permit deference to stand in the way of the relief sought.
It would be naive not to recognize that some prejudice, bias and fear lurks in all of us. But to permit a conviction to be urged based upon such factors or to permit a conviction to stand having utilized such factors diminishes our fundamental constitutional rights.
Furthermore, the prosecution has resources unavailable to the average criminal defendant. Therefore, it is imperative that information which is essential to the defense in the hands of the prosecution be made available to the accused. If trials are indeed searches for the truth rather than efforts to conceal it, full and fair disclosure is necessary to protect and preserve the rights of the accused against the awesome power of the accusor.
Although extended appeals in criminal matters have been widely criticized, the need for review is amply demonstrated by this matter. There is a substantial danger that our society, concerned about the growth of crime, will retreat from the safeguards and rights accorded to the accused by the constitution. The need to combat crime should never be utilized to justify an erosion of our fundamental guarantees. Indeed, the growing volume of criminal cases should make us even more vigilant; the greater the quantity the greater the risk to the quality of justice.
Notwithstanding that the courts which have reviewed this matter have differed, the myriad decisions reflect that petitioners' case has received the most thorough judicial examination our system of justice has to offer. The record is one of which the judiciary can be proud and one from which the public can take comfort; that no matter how complex the issues, how voluminous the record or how ancient the cause, the case received thorough and thoughtful attention from all judges who reviewed it.
For the above reasons, the writ is granted to the Petitioners and their counsel are directed to submit an appropriate order forthwith in conformance with this opinion.NOTES
 Ex parte Yerger, 75 U.S. (8 Wall.) 85, 95, 19 L. Ed. 332 (1868) (Chase, C.J.).
 Indicates reference to volume and page number of transcript from 1976 trial. A date plus "H" and a page number, i.e., 5/19/81H131, connotes transcript date and page number from the trial court's 1981 hearing on remand. PB refers to Petitioners' brief. RB refers to Respondent's brief. PRB refers to Petitioners' reply brief.
 The New Jersey Supreme Court ruled that the summation constituted proper comment, and the petitioners had not established a case of plain error, the applicable standard for such a review by that court. Carter IV, at 107, 449 A.2d 1280.
 The state argues that even if there were minimal prejudicial impact, it was somehow blunted by the great pains taken by the court to minimize jury prejudice through extensive voir dire of the racial attitudes of potential jurors (thereby forewarning the jury that sensitive racial issues would be coming before it). (RB at 124). Even a careful voir dire cannot ensure that in the context of a long and complex trial jurors will not succumb to an appeal based on racial prejudice. Petitioners allege, for example, that the prosecutor's seemingly unrelated reference to the Irish and Greeks may have been appeals to two jurors who immigrated to the United States from those countries. (PB at 109)
 What the state did offer was a transcript of an October 11, 1966 interview with its key witness, Alfred Bello, in which Bello suggested that revenge was the motive. (21T59-61) Bello also testified that he had heard one of the wounded Lafayette Bar patrons say that there had been robbery (22T30). The only other evidence offered indicating police knowledge of a racial revenge motive was the testimony of a Paterson police detective who testified under cross-examination that he was present during the funeral of Holloway in his capacity as an investigator in the Lafayette Bar murders.
 This theory was prohibited at a 1967 pretrial Miranda hearing, the court ruling its prejudicial value outweighed its probative effect. The ruling was reversed for the 1976 trial (PRB at 39, n. 1)
 In its 1982 analysis, the New Jersey Supreme Court utilized the Agurs "might have affected the outcome test," but noted that test was not translatable into the "mere possibility" that the undisclosed information might have helped the defense. "There must be a real possibility that the evidence would have affected the result." Carter IV, 91 N.J. at 113, 449 A.2d 1280. The court then cited United States v. Dansker, 449 F. Supp. 1057 (D.N.J.1977), remanded for evidentiary hearing, 565 F.2d 1262 (3d Cir.1977), cert. dismissed, 434 U.S. 1052, 98 S. Ct. 905, 54 L. Ed. 2d 805 (1978), to suggest that under the "might have affected the outcome test" the defendants must show more than a reasonable likelihood that the evidence could have changed the jury's verdict. Id.