Summary
reversing order granting writ on grounds of no prejudice; while counsel's representation was assumed deficient, defendant received fair trial "even if his defense was flawed"
Summary of this case from McAleese v. MazurkiewiczOpinion
No. 85-1223.
Argued October 15, 1985.
Decided February 4, 1986.
John L. Lachall (Argued), Lachall, Brion, Cohen Lawson, West Chester, Pa., for appellant.
Stuart Suss (Argued), Dist. Atty., West Chester, Pa., for appellee.
Appeal from the United States District Court for the Eastern District of Pennsylvania.
OPINION OF THE COURT
In this habeas corpus petition, we must consider whether the district court, in denying a defendant's claim of ineffective assistance of counsel, correctly applied the standard set out in Strickland v. Washington, 466 U.S. 668, 104 S.Ct. 2052, 80 L.Ed.2d 674 (1984). We conclude that the court did not properly apply Strickland, but that it nonetheless reached the appropriate result because the alleged errors of counsel did not prejudice the defendant. Accordingly, we will affirm.
I.
The long procedural history of this case begins with the 1974 murder conviction of Robert A. McNeil by a Chester County jury. He was accused of having shot John Walker on January 14, 1972, after a hostile exchange of words in Buddy's Apartment Bar in West Chester, Pennsylvania. The evidence showed that, even before the night of the shooting, there had been ill will between Walker and McNeil. A month earlier, Walter Morris, a mutual acquaintance of the two, was talking with McNeil at the Star Social Club when Walker came up and asked Morris, "What are you doing with him"? App. at 537. When Morris asked why Walker did not like McNeil, Walker replied, "The m____ thinks he is cute, if I ever catch him f____ with any of my women, I'm going to kill his black ass." App. at 538.
On the night of the shooting, Walker approached McNeil and two female friends as they sat at the bar. Walker put his hands on McNeil's shoulders and, when McNeil asked why, Walker said he wanted to talk to one of the women. McNeil then told Walker "just to leave me alone and don't put [your] hands on me any more because I don't like anybody touching me." App. at 559. He got up and put on his coat. As McNeil rose, Walker apologized several times, putting his hands on McNeil as he did so.
McNeil then drew a gun from his right pocket, and turned around. Walker backed up, his hands in the air, still apologizing. The barmaid stepped in between them, apparently to try to stop McNeil from shooting, but then got out of the way. App. at 604. As Walker retreated, McNeil fired several shots in rapid succession, killing Walker and wounding a woman bystander. McNeil then left the bar.
After the jury found McNeil guilty of first degree murder, the trial judge sentenced him to life in prison. Following the denial of his post-trial motions, McNeil, who had been represented at trial by privately retained counsel, filed a direct appeal to the Pennsylvania Supreme Court, with the aid of the Public Defender's office. That court rejected McNeil's claims of errors in the trial court's charge to the jury, because trial counsel had failed to object to the charge. However, the case was remanded for further proceedings to address petitioner's claims of ineffective assistance of counsel. Commonwealth v. McNeil, 461 Pa. 709, 337 A.2d 840 (1975).
Represented by still other counsel, McNeil brought an ineffective assistance of counsel claim under the state's Post-Conviction Hearing Act (PCHA), 19 Pa.Cons. Stat.Ann. § 1180 et seq. (Purdon 1977). The court denied relief, holding that the ineffectiveness claims had been waived because McNeil had not raised them on direct appeal, and had failed to charge his appellate counsel with ineffective assistance. The ruling was affirmed by the Pennsylvania Supreme Court. Commonwealth v. McNeil, 479 Pa. 382, 388 A.2d 707 (1978).
McNeil then filed a second PCHA petition, this time alleging that all of his prior attorneys had been ineffective. The PCHA court again denied relief. On appeal, a divided Pennsylvania Supreme Court, now considering the merits of the ineffectiveness claims, affirmed the denial. McNeil then filed a petition for habeas corpus in the District Court for the Eastern District of Pennsylvania.
The district court issued a memorandum and order on December 20, 1983, 576 F. Supp. 1343 (1983) granting the writ. "Viewing the record in its entirety," the district court was "left with the firm conviction that the performance of petitioner's trial counsel fell far short of the quality of representation to which the petitioner was constitutionally entitled." App. at 13-14.
Specifically, the district court faulted trial counsel for failing to inform the jury that Walker, the victim, had previously been convicted of aggravated assault. On the facts of the case, the court reasoned, McNeil could have sought an acquittal based on self defense, or a lesser conviction for voluntary manslaughter, based on a genuine but unreasonable fear for his own life. In either case, evidence of Walker's inclination to violence was critical, and trial counsel's failure to introduce it was "manifestly a serious dereliction." App. at 10. At the PCHA hearing, trial counsel explained his action by stating that since he had introduced witnesses who testified that Walker had a reputation for violence in the community, "I thought that had been covered." App. at 121.
The second error found by the district judge concerned the trial court's charge to the jury. McNeil's counsel did not submit any points for charge, and requested only a "general exception" to the charge as delivered. The trial judge's charge included a long definition of first- and second-degree murder; a charge on the subject of self-defense with stress on the obligation to retreat; and a discussion of that form of voluntary manslaughter based on "heat of passion" and "provocation." However, there was only a brief reference to the theory of voluntary manslaughter based on the defendant's genuine but unreasonable belief that the shooting was necessary in self-defense — the type of voluntary manslaughter the district court viewed as most plausibly applicable to McNeil's defense.
After deliberating for two hours, the jury requested further instructions, asking for "the legal definition of voluntary manslaughter." In response, the trial judge gave an abbreviated definition of voluntary manslaughter, but omitted reference to manslaughter based on genuine but unreasonable apprehension of danger. McNeil's lawyer did not object, nor did he ask that the charge be supplemented. The net effect of all this, the district court concluded, was to undermine McNeil's case that he had killed Walker in fear, whether reasonable or unreasonable. After the writ of habeas corpus was granted, the Commonwealth appealed to this Court.
A short time after the district court's decision, the Supreme Court decided Strickland v. Washington, 466 U.S. 668, 104 S.Ct. 2052, 80 L.Ed.2d 674 (1984), which set out a standard for judging claims of ineffective assistance of counsel under the Sixth Amendment. Upon motion of the Commonwealth, this Court remanded the case to the district court for further proceedings in light of Strickland. On April 3, 1985, the district court issued a new memorandum and order, vacating its previous order "under the compulsion of Strickland" and denying the writ.
In the second order, the district court reaffirmed its previous finding that McNeil's trial counsel had performed deficiently. Along with the actions cited as errors in his 1983 order, the court also noted as error trial counsel's decision to introduce evidence that McNeil had previously been convicted of aggravated battery in the mistaken belief that the then 11-year-old conviction could be used to impeach McNeil's credibility. App. at 755.
Nonetheless, the district court denied the writ because, it held, McNeil had not met his burden for demonstrating prejudice under Strickland. Under that case, "petitioner must shoulder the (virtually impossible) burden of establishing that, if properly presented, his defenses might well have persuaded the jury." App. at 756. The district court continued:
Apparently, therefore, this court is expected to review the trial record and arrive at a conclusion as to whether this particular defendant, in this particular case, might well have achieved a more favorable result at trial if his representation had been constitutionally adequate. In all candor, I find it impossible to express any confident conclusion on that subject . . .
Since the burden is now upon the petitioner, see Strickland, supra, to establish the requisite kind of prejudice, I conclude that my inability to reconstruct what might have happened if petitioner had been represented at trial by adequate counsel means that petitioner has not sustained his burden of proof.
Id. McNeil appealed the denial of the writ of habeas corpus to this Court.
II. A.
In interpreting the Sixth Amendment's guarantee of the right to counsel, the Supreme Court has recognized that "the right to counsel is the right to the effective assistance of counsel." McMann v. Richardson, 397 U.S. 759, 771 n. 14, 90 S.Ct. 1441, 1449 n. 14, 25 L.Ed.2d 763 (1970). Until 1984, however, when it decided Strickland, the Supreme Court had not set out a standard for judging ineffective assistance claims premised upon allegations of incompetent representation, thus leaving the standard to be determined by the various federal and state courts.
In this Circuit, a two-part test had evolved. To obtain a reversal on an ineffective assistance claim under the Sixth Amendment, a defendant was required to demonstrate first that his lawyer's performance fell short of "the customary skill and knowledge which normally prevails at the time and place." Moore v. United States, 432 F.2d 730, 736 (3d Cir. 1970) (in banc). The Moore court, however, did not insist on a separate showing of prejudice; rather it stated that the question whether the defendant was prejudiced by counsel's errors was "evidentiary" on the issue whether the attorney's performance fell below a reasonable standard of competence. Id. at 137. Soon afterward, however, this Court made clear that a separate showing of prejudice was required. See, e.g., United States ex rel. Green v. Rundle, 434 F.2d 1112, 1113 (3d Cir. 1970).
Sitting in banc, this Court in United States v. Baynes, 687 F.2d 659 (3d Cir. 1982), elaborated on the content of the second prong of an ineffectiveness claim, the prejudice showing. "[A] court," it stated, "should expect the habeas petitioner to demonstrate that there is a `reasonable possibility' that had the error of which he complains not occurred, the jury might have arrived at a different conclusion." Id. at 670. In conducting the prejudice inquiry, we declared, a reviewing court should be guided by the harmless error standard of Chapman v. California, 386 U.S. 18, 87 S.Ct. 824, 17 L.Ed.2d 705 (1967), which "permits a finding of harmless error only where it is concluded beyond a reasonable doubt that no prejudice resulted from the identified constitutional violation." Baynes, 687 F.2d at 670-71. Thus, "any substantial doubt must be resolved in favor of the defendant." Id. at 670.
Two years later, the Supreme Court for the first time addressed the issue of an ineffectiveness standard when it decided Strickland v. Washington, 466 U.S. 668, 104 S.Ct. 2052, 80 L.Ed.2d 674 (1984). While elaborating on existing law and modifying it somewhat, Justice O'Connor's opinion suggests that the Court intended no dramatic change in the standard being applied in the lower federal and state courts. Noting that all the federal courts of appeals had adopted substantially the same standard for attorney performance, the Supreme Court declined to depart from such a standard, holding, "The proper measure of attorney performance remains simply reasonableness under prevailing professional norms." Id., 104 S.Ct. at 2065. In examining counsel's performance, Justice O'Connor added, a court must judge "the reasonableness of counsel's challenged conduct on the facts of the particular case, viewed as of the time of counsel's conduct," with a recognition that "counsel is strongly presumed to have rendered adequate assistance and made all significant decisions in the exercise of reasonable professional judgment." Id. at 2066.
As to the prejudice inquiry, the Supreme Court adopted a test similar to that being applied in this Circuit, although significant, it did not incorporate the Chapman standard. Strickland laid down the following test:
The defendant must show that there is a reasonable probability that, but for counsel's unprofessional errors, the result of the proceeding would have been different. A reasonable probability is a probability sufficient to undermine confidence in the outcome.
Id. at 2068.
Further insight into the meaning of this standard may be derived from examining the alternatives rejected. Strickland first set aside a lenient standard that would allow a defendant to prevail on a showing that the errors "had some conceivable effect on the outcome of the proceeding," or "impaired the presentation of the defense." At the same time, however, it declined to adopt the strict "outcome determinative" standard of United States v. DeCoster, 624 F.2d 196, 208, 211-12 (D.C.Cir.) (in banc) (plurality opinion of Leventhal, J.), cert. denied, 444 U.S. 944, 100 S.Ct. 302, 62 L.Ed.2d 311 (1979). Under the latter test, a defendant would have to "show that counsel's deficient conduct more likely than not altered the outcome in the case." Strickland, 104 S.Ct. at 2068. Although the "outcome determinative" standard "reflects the profound importance of finality" in criminal proceedings, "the result of a proceeding can be rendered unreliable, and hence the proceeding itself unfair, even if the errors of counsel cannot be shown by a preponderance of the evidence to have determined the outcome." Id.
While setting out particular standards for judging performance and prejudice, Strickland also cautioned that the tests were not designed as "mechanical rules," and prescribed a broader rule governing the resolution of ineffectiveness claims:
Although those principles should guide the process of decision, the ultimate focus of inquiry must be on the fundamental fairness of the proceeding, whose result is being challenged. In every case the court should be concerned with whether, despite the strong presumption of reliability, the result of the particular proceeding is unreliable because of a breakdown in the adversarial process that our system counts on to produce just results.
Id. at 2069.
Strickland, then, alters somewhat the ineffectiveness test applied previously in this Circuit, but it does not significantly modify the method of inquiry required of reviewing courts. Under the Supreme Court's test, the prejudice standard appears somewhat more onerous for defendants, since the Court does not use this Circuit's rule that a court determining that no prejudice resulted from counsel's errors must make this finding beyond a reasonable doubt. Moreover, Strickland stresses the "strong presumption of reliability" to be afforded the results of the adversarial process. At the same time, however, the "reasonable probability" test of Strickland appears quite similar to the "reasonable possibility" test of Baynes. And the method of inquiry prescribed in Strickland is not inconsistent with the requirement of Baynes that a court undertake "a careful inquiry into the particular circumstances surrounding each case." 687 F.2d at 665. Nor does Strickland change the fact that "[a] determination whether any given actions or omission by counsel amounted to ineffective assistance cannot be divorced from consideration of the peculiar facts and circumstances that influenced counsel's judgment. In this fact-laden atmosphere, categorized rules are not appropriate." Baynes, 687 F.2d at 665, quoting United States v. DeCoster, 624 F.2d at 203.
The dissent interprets Strickland to state that its test does not impose upon a habeas petitioner a heavier burden of proving prejudice than previously applied in this Circuit. Post at 456. A fair reading of Strickland, however, does not support this view. The dissent relies on two sentences in Justice O'Connor's opinion: "With regard to the prejudice inquiry, only the strict outcome-determinative test, among the standards articulated in the lower courts, imposes a heavier burden on defendants than the tests laid down today. The difference, however, should alter the merit of an ineffectiveness claim only in the rarest case." Strickland, 104 S.Ct. at 2069.
Although the Court's wording is not without ambiguity, it is clear that the first sentence means that of the standards applied in the lower courts, only the D.C. Circuit's "outcome determinative" test imposes a greater burden of showing prejudice that the Strickland standard; it follows logically that the other standards impose either as great a burden or a lesser burden than Strickland. Thus, this sentence does not, as the dissent insists, "explicitly" state that Strickland imposes the same standard as this Circuit, post at 456, and other language in the opinion prescribing a "presumption of reliability" makes clear that the Strickland standard is somewhat more onerous for defendants than our Baynes standard.
The second sentence is less clear. The principal question is what "difference" the Supreme Court is referring to. Because the previous sentence focused primarily on the D.C. Circuit's "outcome determinative" standard, it seems most likely that the "difference" refers to the disparity between the D.C. Circuit test and the Strickland formulation. Thus, the Supreme Court may be suggesting that its result will likely accomplish the same effect in most cases as the "outcome determinative" test. Alternatively, the Supreme Court may intend to suggest more broadly that the difference in the exact wording of prejudice and performance tests is less significant than an overall consideration of the "fundamental fairness" of the proceeding under review. Either interpretation is consistent with our application of Strickland in this case.
B.
The foregoing discussion would appear to demonstrate that the district court in this case misapplied Strickland. More important, the court erred by not undertaking a careful examination of the record in an attempt to determine whether the mistakes it found counsel had committed had prejudiced McNeil within the meaning of Strickland. The district court characterized the defendant's burden under the prejudice prong of Strickland as the "virtually impossible" task of "establishing that, if properly presented, his defenses might well have persuaded the jury." App. at 756. The district court noted that on the basis of statistics, few defendants who go to trial are acquitted. Moreover, it asserted that while Strickland appears to envision a close evaluation of the record to determine the likelihood of a different outcome absent the errors by counsel, it "[found] it impossible to express any confident conclusion on that subject."
However, although it imposes a substantial burden for showing prejudice, Strickland does not establish a "virtually impossible" burden. Indeed, it relaxes the standard previously employed by some courts, such as the District of Columbia Court of Appeals. Strickland requires a searching inquiry into whether an attorney's alleged errors have prejudiced his client. Such a retrospective inquiry into the facts and circumstances of a criminal trial may not be easy, yet it is the same type of investigation courts frequently conduct when they consider the materiality of "exculpatory information not disclosed to the defense by the prosecution, United States v. Agurs, 427 U.S. at 97, 104, 112-13 [96 S.Ct. at 2392, 2397, 2401-02, 49 L.Ed.2d 342 (1976)] and [when they apply] the test for materiality of testimony made unavailable to the defense by Government deportation of a witness. United States v. Valenzuela-Bernal, 458 U.S. [858] at 872-74 [102 S.Ct. 3440 at 3449-50, 73 L.Ed.2d 1193 (1982)]." Strickland, 104 S.Ct. at 2068. Thus, "[t]he prejudice standard [of Strickland] does not erect an insurmountable obstacle to meritorious claims, but rather requires courts carefully to examine trial records in light of both the nature and seriousness of counsel's errors and their effect in the particular circumstances of the case." Id. at 2073 (Brennan, J., concurring in part). If defendants' Sixth Amendment right to the effective assistance of counsel is to have any meaning, courts must apply these standards conscientiously and carefully.
III.
Turning to the facts of this case, we must now apply Strickland and determine whether the alleged errors of counsel at McNeil's trial rose to the level of ineffective assistance of counsel sufficient to abridge the Sixth Amendment. Strickland encourages courts to resolve cases wherever possible on grounds of prejudice, so that ineffective assistance claims do not become unduly burdensome to defense counsel. Id. at 2069-70. Accordingly, assuming that the alleged errors of trial counsel do constitute deficient performance, we must determine whether "there is a reasonable probability that, but for counsel's unprofessional errors, the result of the proceeding would have been different." Id. at 2068.
In making this determination, we review the record on a plenary basis, for "the performance and prejudice components of the ineffectiveness inquiry are mixed questions of law and fact" to which the requirement of deference contained in 28 U.S.C. § 2254(d) (1982), does not apply. Strickland, 104 S.Ct. at 2070.
First, absent the alleged errors of counsel cited by the district court, McNeil's prior conviction 11 years earlier for aggravated battery, brought out by his attorney on direct examination, would not have been placed before the jury, although his prior conviction for larceny of an automobile would have been. Second, the jury would have been informed of the prior conviction of the victim for aggravated assault. At trial, defense counsel failed to bring out this fact. The prior conviction of Walker arguably could have undermined the credibility of the witness who testified as to the victim's peaceful nature. And of greater import, the jurors would have had this item of testimony before them as evidence of Walker's propensity for violence when they deliberated on whether McNeil shot him out of fear, rather than by premeditation, as required for first-degree murder. Evidence of Walker's aggressive nature would have been relevant either to a self-defense verdict for McNeil, which excuses a shooting perpetrated in reasonable fear of one's own life, or a voluntary manslaughter verdict, which applies to shootings in "unreasonable fear."
Finally, absent the failure by McNeil's counsel to object to the incomplete supplemental instruction, the jurors would have been reminded of the theory of voluntary manslaughter applicable to a defendant who has killed someone because of an unreasonable belief that his life is in jeopardy. When the jurors returned after two hours of deliberation to ask for the definition of voluntary manslaughter, a more alert counsel would have recommended that this element of the defense be included in the definition; the judge undoubtedly would have included it, since it was set forth in his initial charge.
Had none of the above errors or omissions occurred, McNeil maintains, the chances of his being convicted of first-degree murder would have been greatly reduced. Instead of retiring with evidence that McNeil had a record of violent crime and Walker did not, the jury would have been in the opposite position, with evidence of Walker's violent criminal tendencies and much less evidence of McNeil's — at worst, evidence that McNeil had once stolen a car. Moreover, the argument continues, the jury would have had its attention drawn more sharply to the availability of a verdict consistent with this evidence: a voluntary manslaughter conviction, on the theory that McNeil shot Walker out of unreasonable fear for his own life. All these factors, it is argued, create a reasonable probability that the outcome would have been different without the errors, for "the errors committed here had a pervasive effect on the inference to be drawn from the evidence, altering the entire evidentiary picture." Strickland, 104 S.Ct. at 2069.
On balance, however, we do not think these errors had a sufficiently great impact. None of them altered the crucial factual finding that McNeil shot a man who was backing away from him with his hands up and apologizing. It is true that Walker grabbed McNeil from behind when he first approached him at the bar, that the two exchanged words, and that McNeil testified Walker threatened him during the exchange. No other witness heard this conversation, however, while witnesses heard McNeil's angry warning to Walker to keep his hands off. Moreover, after this, there is no evidence that Walker made any menacing gestures, and overwhelming evidence that he sought to calm McNeil as he rose to get his coat by putting his hands on his chest or shoulders and saying that he was sorry or that it was "all right." App. at 205-06; 302-03; 336; 380. It is undisputed that Walker was apologizing to McNeil in the moments immediately prior to the shooting. Evidence of the victim's prior violence and reduced evidence of the defendant's criminal past, then, do not undercut the impression that McNeil purposefully shot Walker without any fear of the victim at the time he shot.
McNeil also testified that Walker had taken a hat off his head and shoved it into his face. None of the other witnesses, however, supported this account. App. at 579.
Indeed, a fair reading of McNeil's own trial testimony would largely eliminate fear as his dominant emotion at the time of the shooting. In describing how he felt as he was putting on his coat, McNeil testified, "I was getting ready to go. He [Walker] just kept badgering at me . . . He just kept pulling and just picking at me like I was some kind of little child or kid or something." App. at 560. McNeil did not say he suspected an attack. When he turned his back to Walker and felt his hand on his back, McNeil testified, "I don't know what he was going to do." When specifically asked whether he was fearful while Walker was backing away with his hands up saying, "I'm sorry," McNeil replied only, "I remember seeing the mass in front of me. That's the last thing I remember seeing of Johnny." App. at 585-585.
Further, McNeil's immediate response to Walker's approach at the bar is far more consonant with an attitude of aggressive self-confidence than fear, as also suggested by Walker's quick apology, and the effort by the barmaid to stop McNeil. Thus, the lack of a supplemental instruction on "unreasonable fear" would not be a crucial omission, for there was virtually no evidence to support such a verdict, while there was overwhelming evidence of a purposeful killing. "[A] verdict or conclusion only weakly supported by the record is more likely to have been affected by errors than one with overwhelming support." Strickland, 104 S.Ct. at 2069. This case falls into the latter category.
IV.
We do not mean to express approbation of the representation McNeil received from his attorney in this case. To the contrary, it is as clear to us as to our dissenting colleague that a better prepared, more adept lawyer would have represented McNeil far more vigorously, and would not have made some of the dubious decisions detailed earlier. Upon reviewing the cumulative effect of these actions and omissions, however, we do not think there is a "reasonable probability" that without them, the result of the trial would have been different. It is possible that the outcome would have been different, of course, but the possibility is not "sufficient to undermine confidence in the outcome." Id. at 2068.
More broadly, we believe the record shows that, even if his defense was flawed, McNeil received a fair trial. At trial, McNeil put on defense witnesses who presented evidence favorable to his case. He testified at length, providing the jury with his account of the events in question. Indeed, his ineffective assistance argument is premised upon the proposition that the jury had before it sufficient evidence to find him guilty of a charge less than first degree. The jury, moreover, was instructed on lesser degrees of guilt.
Under these circumstances, we do not believe there was "a breakdown in the adversarial process," nor do we have any significant doubt as to the reliability of the outcome. As the Supreme Court has noted in similar contexts, a "`defendant is entitled to a fair trial but not a perfect one.'" Bruton v. United States, 391 U.S. 123, 135, 88 S.Ct. 1620, 1627, 20 L.Ed.2d 476 (1968), quoting Lutwak v. United States, 344 U.S. 604, 619, 73 S.Ct. 481, 490, 97 L.Ed. 593 (1953). Because McNeil received that here, the judgment of the district court will be affirmed.
The question for decision is not whether McNeil was prejudiced "even if his counsel did err." The district court, the majority here, and I all agree that McNeil's legal representation fell below any acceptable minimum of competency. In plain talk, we all agree that McNeil's lawyer committed malpractice. It is even difficult to see how McNeil would have been worse off if the trial court had permitted a non-lawyer to represent him. I am unable to agree, therefore, with the majority's views that, first, the jury would surely have found McNeil guilty of murder absent the errors of his attorney, and second, that the Strickland standard for prejudice differs in meaningful degree from the one formerly applied in this circuit.
Had the court denied McNeil a lawyer, of course, the grant of habeas corpus would have been automatic. Gideon v. Wainwright, 372 U.S. 335, 338-45, 83 S.Ct. 792, 793-97, 9 L.Ed.2d 799 (1962).
The majority has summarized the errors of McNeil's attorney, but it is useful to recapitulate in order to bring certain details to light. McNeil's counsel not only unnecessarily elicited the criminal record of his client, he did not even object when the trial judge instructed the jury that they could use that record to judge McNeil's credibility. And that criminal record included a twelve year old conviction for aggravated assault. It is surely clear that McNeil's credibility was critical to his partial defense of voluntary manslaughter. The entire issue before the jury was the sincerity of his claim that he was in fear when he shot Walker. His attorney voluntarily impeached McNeil's character and credibility in ways so improper that were they done by the prosecution, there would have been a certain mistrial or reversal. So much for the jury's view of the character of the accused.
Voluntary manslaughter differs from homicide only in that the fact-finder determines the killer believed "the circumstances to be such that, if they existed, would justify the killing . . . but his belief is unreasonable." Pa.Cons.Stats. Ann. § 2503(b) (Purdon 1983).
And what of the character and reputation of the deceased? Counsel for the defendant did not even attempt to bring forth the clearly proper evidence that Walker had been convicted of assault with intent to kill. McNeil's attorney was content to rely on the bare word of his client that he "believed" Walker to be dangerous. Not even the appearance of witnesses, called by the prosecution, and testifying to the deceased's "good reputation as a peaceful and non-violent person," could stimulate McNeil's lawyer to bring forth Walker's conviction for assault with intent to kill. The jury's view of the character of the two actors in this drama was therefore a complete distortion of what it would have been had McNeil been competently represented. Had McNeil been properly represented, the jury would have heard only that Walker had previously tried to kill someone and nothing of McNeil's assault conviction. As it was, McNeil could hardly have been worse off with no lawyer at all. Indeed, had McNeil waived counsel under Faretta v. California, 422 U.S. 806, 95 S.Ct. 2525, 45 L.Ed.2d 562 (1975), no district judge would have sat silent and permitted the prosecutor to do to an unrepresented defendant what McNeil's defense attorney did to McNeil.
App. at 13 (Judge Fullam's Dec. 20, 1983 opinion accompanying the grant of habeas corpus).
The majority holds that none of this matters because it does not have "any significant doubt concerning the reliability of the outcome" of the trial. Ante, at 451. The evidence of Walker's violence, and the omission of the evidence of McNeil's assault conviction would not, according to the majority, "undercut the impression that McNeil purposefully shot Walker without any fear of the victim at the time he shot." Ante at 451.
There were actually two episodes of physical contact — one before McNeil put on his coat and one after. On the first, eye-witness Lois Purnsley testified, "Mr. Walker came up and sort of like grabbed Mr. McNeil but when he grabbed him it knocked Mr. McNeil into me and I almost fell off my stool." App. at 516. Eye-witness Samuel Dorsey, testifying for the Commonwealth, stated, "[Walker] went over and shook McNeil, put his hands on him, shook him." App. at 379. Eye-witness and Commonwealth witness Deborah Anderson testified that Walker "grabbed" McNeil when McNeil's back was turned. App. at 301. When asked if Walker grabbed McNeil with such force that it turned McNeil around, Anderson responded, "I'm not sure." App. at 301.
The majority's version of the second contact is that after McNeil put his coat on to leave, "Walker apologized several times, putting his hands on McNeil as he did so." Ante, at 445. Lois Purnsley reported: "Mr. Walker was like in back of him and he was walking and kept, you know, sort of nudging him on the shoulder, you know, kept saying something to him as he was going to the door." App. at 517. Deborah Anderson testified that she did not know if Walker grabbed McNeil or how hard he grabbed him, App. at 303, but she agreed that Walker "spun [McNeil] around or put his hand on his shoulder." App. at 305.
The majority makes only fleeting allusion to the fact that after the first contact, Walker and McNeil exchanged words for "maybe a minute or two," according to eyewitness Patricia Gonzalez. App. at 237. Every eyewitness testified that they could not hear what was said. App. at 205 (Elaine Smith); 237 (Gonzalez); 269-70, 280 (Bernadette Bolden); 304-06 (Anderson); 325, 327, 337-38 (Georgia Harvey); 358-59 (Sandra Davis); 379, 389 (Dorsey); 519 (Purnsley). McNeil's uncontradicted account was that "[Walker] said to me that nigger, I ought to kill you. Then he tried to bump me or push me in another manner and almost knocked me on the bar. . . . I was just getting ready to go. He just kept badgering me. . . ." App. at 559. McNeil continued:
He just kept pulling and picking at me like I was some kind of a little child or kid or something. . . . [H]e had put his hand on me again to try to turn me — I don't know whether he was going to grab me again, push me or what he was going to do but I felt him touch me again, to pull me kind of around. I don't know what he was going to do. He already done threatened me so I reached in my pocket and I just pulled out the gun and just kept firing.
App. at 559-60. The majority treats McNeil's remark "I don't know what he was going to do" as an admission that McNeil did not suspect an attack and as evidence that fear was not his dominant emotion. Ante, at 451. But when the remark is read in context, it appears to be an expression of fear over what Walker would do next.
McNeil also recounted four previous occasions when Walker had taunted him and threatened his life, App. at 562-66, 572-76, one of which was corroborated by Walter Morris, App. at 538, as the majority points out. Ante, at 445. Walter Morris and Alice Carcasia testified to Walker's bad reputation for violence and to his dislike for McNeil. App. at 539-40, 546-49. Morris reported that Walker carried a pistol or knife on previous occasions. App. at 539-40. McNeil said he knew Walker carried a weapon and had shot a woman. App. at 582-83.
The point is not merely that the majority's version of the facts is wrong but that the jury could very well have formed an impression of the circumstances that would have justified a voluntary manslaughter verdict. The majority's conviction that "there was overwhelming evidence of a purposeful killing," ante at 451, is irrelevant because, like homicide, voluntary manslaughter is defined as purposeful killing. The conduct of the jury's deliberations certainly shows they were more troubled by McNeil's defense, even with all of its errors, than the majority thinks possible.
Unreasonable fear of harm does not excuse a killing, but it does reduce the degree to voluntary manslaughter. Commonwealth v. McNeil, 497 Pa. 187, 439 A.2d 664, 669 (Pa. 1981). The relevant type of voluntary manslaughter is defined in the statute as follows: "A person who intentionally or knowingly kills an individual commits voluntary manslaughter if at the time of the killing he believes the circumstances to be such that, if they existed, would justify the killing . . . but his belief is unreasonable." Pa. Cons.Stat.Ann. § 2503(b) (Purdon 1983) (emphasis added). McNeil denied killing Walker intentionally, App. at 581-82, but the jury could have returned a voluntary manslaughter verdict even if it found the killing was intentional. The fact that Walker was retreating when shot also does not preclude a voluntary manslaughter verdict. See Commonwealth v. Walley, 466 Pa. 363, 353 A.2d 396, 399-400 (Pa. 1976); Commonwealth v. Edwards, 466 Pa. 336, 353 A.2d 383, 384 (Pa. 1976).
McNeil's lawyer next made a series of mistakes in connection with the jury charge. He submitted no requests to charge. The trial judge's charge to the jury fills forty-six typescript pages, App. at 604-50, of which five pages are devoted to murder, App. at 616-21, 639; five are devoted to a largely negative discussion of self-defense, App. at 634-39; three are devoted to the irrelevant form of manslaughter based on heat of passion, App. at 621-23, 639-40; and only two sentences pertain to the form of voluntary manslaughter based on unreasonable fear:
If [the killing] was committed under the influence of an uncontrollable fear of either death or great bodily harm caused by the circumstances but without the presence of all the ingredients necessary to excuse the act on the ground of self-defense, the killing is manslaughter and not murder. Remember that the uncontrollable fear of death or great bodily harm, conceivable as existing, but not necessarily justified by the immediate circumstances, the killing is voluntary manslaughter.
App. at 640. This explanation is simply tacked onto the discussion of heat-of-passion manslaughter in a way that could have led the jurors to think it was elaboration of heat-of-passion manslaughter. App. at 640.
After two hours of deliberation, the jury was sufficiently concerned about McNeil's defense to emerge with a question: "Would you be so kind as to give us the following: What is the legal definition of voluntary manslaughter? Factors necessary for such." App. at 651. In response, the trial judge gave an instruction on heat-of-passion manslaughter that omitted any reference to the kind of manslaughter relevant to the defense evidence. App. at 651-52. Counsel for McNeil failed even to object. The majority suggests that this was a mere failure to be "alert." I believe that finding understates what amounts to yet another significant act in what was a gross failure of representation.
On the facts of this case, the jurors could very well have believed that McNeil was in genuine, if unreasonable, fear of his life or bodily harm, because there was much evidence that Walker's conduct was threatening. The jury's request for a repeat of the charge on voluntary manslaughter shows that such a verdict was under consideration even though the trial judge's only previous explanation of the relevant form of voluntary manslaughter was virtually smothered in the rest of his instruction.
The jury went back to work. After the jury requested and received a supplementary charge on murder, McNeil's attorney apparently (the sidebar discussion is not transcribed) requested a further charge on voluntary manslaughter, App. at 656, but the trial judge refused. Counsel's failure to make a record or to object did not preserve the ruling for appeal, and, as a result, McNeil could not even bring the matter to the attention of the appellate courts of the Commonwealth.
The majority concedes that McNeil was the victim of the ineffective assistance of his counsel, but is nevertheless confident that McNeil received a "fair trial." Because I am unable to agree, I am constrained to respectfully dissent.
Perhaps that is all that needs to be said, yet I think another word may be timely. The unstated difficulty beneath all these "ineffective assistance" of counsel cases is the dilemma of fairness versus finality. Courts can compel the state to provide counsel, but they have not yet even attempted to guarantee the competency of the counsel that they provide. And once the trial is over, considerations of finality give us an institutional reason to do all we can to avoid trying the case again.
It is tempting to say, for example, McNeil picked his own trial counsel and should bear the risk of an improvident choice. The state, after all, is as much a victim as the defendant when counsel's malpractice results in a retrial, but the state cannot select its adversaries or do anything much to help when it sees an incompetent adversary during trial. Yet merely to state that proposition openly compels us to reject it. McNeil was entitled to rely upon his counsel's license to practice law when he selected his lawyer. The negligence, if any, is not McNeil's.
Ten years ago, sitting on this court by designation, I tried to call attention to the anomaly of courts finding plain error or ineffective assistance of counsel, and then forcing a retrial without doing anything about the unprofessionalism of the attorney:
It is of course a clear anomaly that historically both bench and bar have resisted certification procedures designed to regulate and improve the quality of the trial bar, content to indulge in the fiction that a graduate of a law school who has passed a bar examination is qualified to try any case, from the simplest intersection accident to the most serious criminal indictment. We should not forget that it is the courts themselves who license, and thus extend an implied promise of the competence of the attorney to all. When the plain error rule is invoked, or even discussed, there ought to be coordinate consideration given to remedial action, designed to improve the quality of the representation either in general or particular. The restrictions which the Sixth Amendment and the double jeopardy clause of the Fifth Amendment place upon the powers of the trial courts to intervene in an ongoing trial make this issue of crucial and immediate significance.
United States v. Mitchell, 540 F.2d 1163, 1170 n. (3d Cir. 1976) (STERN, D.J., concurring).
The state has been victimized, to be sure, whenever a retrial is necessary because of a breakdown in the adversarial system. But the state pays the least. The defendant, after all, might well be in jail because his counsel did not perform his job.
I think it bears repeating that if McNeil had been unrepresented, we would inquire no further. There would be no mental re-enactment of the trial with imagined competent counsel to test whether his absence made a difference. The pernicious doctrine of Betts v. Brady, 316 U.S. 455, 62 S.Ct. 1252, 86 L.Ed. 1595 (1942) has been put aside. Gideon v. Wainwright, 372 U.S. 335, 83 S.Ct. 792, 9 L.Ed.2d 799 (1963). Of course, the presence of counsel does make a difference, but it can only make a difference when that person acts like a lawyer. A drunken lawyer, an incapacitated lawyer, an incompetent lawyer is not a lawyer. That is not to say that we must approach both classes of cases, the non-represented and the claimed incompetently represented, in the same way. Without even if the minimal guarantee of fairness provided by the presence of a licensed lawyer, we are properly unwilling even to attempt to rewrite the judicial history of the case. The presence of a fully licensed representative, on the other hand, is enough guarantee of a functioning adversarial process to entitle us to examine claims of his mistakes and errors against consideration of whether those errors made a difference. The spirit in which we do this, however, is the entire issue. It is to Strickland that we must now look.
The district judge whom we review has twice found McNeil to have been severely prejudiced by his lawyer's conduct. The first time, he granted the petition under the Baynes test. This decision was not disturbed. He was only instructed to reconsider his decision under Strickland. On remand, the district judge denied the writ, not because McNeil was unprejudiced, but only because the district judge believed that Strickland had made the grant of any habeas petition alleging ineffective assistance of counsel "virtually impossible."
The majority acknowledges that the district judge "misapplied" Strickland, Ante, at 449, that he mistakenly read Strickland to dramatically increase the burden upon petitioners to show that counsel's mistakes were significant to the result. My difficulty is that the majority now holds that Strickland has indeed made it more difficult for petitioners, and the majority has applied that higher standard to change the outcome of this case. I cannot agree because Strickland specifically says that it does not create a high standard.
The former test in this circuit, under which the district judge initially granted McNeil's habeas petition in 1983, was that the defendant demonstrate a "reasonable possibility" that counsel's errors affected the result. United States v. Baynes, 687 F.2d 659, 670 (3d Cir. 1982). Justice O'Connor did not create a different standard with the "reasonably probability" language. The majority does not seem to claim that "reasonable probability" is a higher standard than "reasonable possibility," but the majority could have been much clearer in rejecting this view. For Justice O'Connor did everything she could to explain that "reasonable probability" does not mean "greater than 50%": ". . . [W]e believe that a defendant need not show that counsel's deficient conduct more likely than not altered the outcome of the case." Strickland, 104 S.Ct. at 2068. "The result of a proceeding," the court added, "can be rendered unreliable, and hence the proceeding itself unfair, even if the errors of counsel cannot be shown by a preponderance of the evidence to have determined the outcome." Id. For its view that the Strickland standard is "somewhat" different, the majority relies on Baynes' suggestion that "any substantial doubt must be resolved in favor of the defendant," 687 F.2d at 670, a phrase that does not appear in Strickland.
But Strickland does not say that any substantial doubt must be resolved in favor of the government. In fact, Strickland holds that to show reasonable probability, a petitioner need not show it more likely than not that counsel's mistakes altered the probable outcome. The Strickland court goes on to say: "A reasonable probability is a probability sufficient to undermine confidence in the outcome." 104 S.Ct. at 2068. This makes the Strickland standard indistinguishable from that of Baynes, because if there were "any substantial doubt" about the outcome ( Baynes), it would surely "undermine confidence in the outcome" (Strickland). Under both tests, substantial doubt inures to the benefit of the petitioner.
The majority's view that there is "somewhat" of a difference between the Baynes and Strickland standards, and the imposition of that difference in this case, ignores a clear injunction of Strickland:
With regard to the prejudice inquiry, only the strict outcome-determinative test, among the standards articulated in the lower courts, imposes a heavier burden on defendants than the tests laid down today. The difference, however, should alter the merit of an ineffectiveness claim only in the rarest case.
Id. at 2069. The Supreme Court plainly sought to preclude courts from doing exactly what the majority has done: From latching onto mere differences in wording and thereby imposing a higher standard on habeas petitioners than they formerly had. The Court explicitly defined its prejudice test as no heavier than the ones in force in most circuits including this one, to expressly avoid the result in this case.
Because I am unable to find this to be "the rarest case," because I believe McNeil was hopelessly prejudiced by the egregious errors of his lawyer, and because I believe the majority has wrongly increased the burdens that future habeas petitioners will face in the district courts of this circuit, I am constrained respectfully to dissent.