OPINION
A jury сonvicted Daniel Bedford of the aggravated murder of Gwen Toepfert and the murder of John Smith, and at the jury’s recommendation a state trial court sentenced him to death. The Ohio courts affirmed his convictions and sentence on direct review and denied postconviction relief. Bedford sought a writ of habeas corpus under 28 U.S.C. § 2254, which the district court denied. We affirm.
I.
In 1978, Bedford met Toepfert, whose father owned the bar where Bedford worked, and for the next several years the two were involved in an “on-again, off-again” relationship. JA 491. By 1984, they were estranged.
See State v. Bedford,
Bedford’s feelings for Toepfert remained, however, prompting him to try to “rekindle [their] prior romance.” Id. On April 21, 1984, he visited her apartment bearing a gift and hoping to make amends — only to learn that Toepfert’s new boyfriend, John Smith, was already there. Id. Three days later, Bedford tried again. At around 2:30 a.m. on Tuesday, April 24, Bedford, who had spent the evening working at one bar and patronizing another, telephoned Toepfert’s apartment — only to learn from her roommate, Jo Ann Funk, that Toepfert was asleep and that Smith was with her. Id.
Later that morning, Funk woke to the sounds of “gunshots and screams.” Id. Apparently overcome by Toepfert’s rejection, Bedford entered her apartment armed with a .38 revolver and a shotgun, shot Smith after a brief struggle and shot Toepfert. During the melee, Toepfert ran into Funk’s bedroom, screaming that she had been shоt. Bedford found her there and shot her again with the revolver and the shotgun. Smith and Toepfert died from the gunshots. See id.
Bedford fled to Tennessee. Once there, he visited an acquaintance, to whom he confessed his crime, and who reported Bedford to the police. After Tennessee police arrested Bedford (and Mirandized him), he gave a statement admitting the crimes and eventually gave a similar statement to Cincinnati authorities. Id.
An Ohio jury convicted Bedford of the aggravated murder of Toepfert and the murder of Smith.
Id.
at 916. After a mitigation hearing, the jury recommended the death penalty, and the trial court agreed.
Id.
On direct review, the state court of appeals and the Ohio Supremе Court affirmed Bedford’s conviction and death sentence.
See State v. Bedford,
No. C-840850,
In 1992, Bedford filed a federal petition for habeas corpus in the district court. As amended, his petition raised 87 separate grounds for relief. In a pair of thorough opinions spanning 251 pages, the district court denied each of Bedford’s claims. Most of the claims, the court concluded, were procedurally defaulted or otherwise not cognizable in federal court, and the remainder failed on the merits. The court granted a certificate of appealability on several claims.
See Slack v. McDaniel,
II.
Because Bedford filed his federal habeas petition before AEDPA’s effective date, AEDPA’s standard of review does not apply,
see Lindh v. Murphy,
A.
Bedford first claims that the trial court unfairly limited his questioning of prospective jurors during voir dire: (1) by too quickly dismissing four prospective jurors for cause that he wished to rehabilitate and (2) by precluding his counsel from asking certain questions of the jurors.
1.
A prospective death-penalty juror may be struck for cause if he is “substantially impaired in his ... ability to impose the death penalty under the state-law framework.”
Uttecht v. Brown,
The four dismissed jurors each expressed views that qualified them as “substantially impaired.” Juror Herweh told the court that he “definitely” did not think he could sign a death-penalty recommendation, JA 2192, even if the aggravating factors outweighed the mitigating factors. Juror Tucker did not think she “could be a part of convicting someone to the death sentence,” would not recommend a death sentence under any circumstance and could “not follow [a] law” requiring her to do so. JA 2132-34. And Jurors Dotterweich and Jordan stated that they could not sign a verdict reсommending the death penalty. Based on these statements, the trial court had ample cause to excuse each juror,
see Dennis,
Bedford counters that his counsel might have rehabilitated the jurors had the trial judge not cut short each colloquy. But the court did allow Bedford’s lawyers to follow up with questions after initial inquiries elicited disqualifying responses, and each time the additional questions confirmed the juror’s unwillingness to sign a death verdict.
The question, then, is not whether the trial court was required to permit followup questions; it is whether the court was required to permit still further follow-up questions. Bedford contends that, had the jurors been reminded that their task re *232 quired them only to make a recommendation to impose the death penalty the jurors might have modified their views. But Bedford’s counsel did mention to all four jurors that they would make only a recommendation.
Bedford adds that further questioning might have shown that the jurors were “simply confused” about the task before them, not unwilling to do their duty. Br. at 112. But chalking up the jurors’ statements to confusion does not help Bedford, because voir dire responses that signal serious confusion about the jury’s role in the process suffice to excuse a juror.
See Morales v. Mitchell,
Even if Bedford could show that the trial court erred in excusing the jurors, at any rate, he still could not obtain relief. To prevail, hе must show not only that the trial court’s decision was incorrect but also that it resulted in an actually biased jury.
Hill v. Brigano,
2.
Also unavailing is Bedford’s claim that the trial court improperly limited the scope of questioning at voir dire. The Constitution “does not dictate a catechism for
voir dire,
but only that the defendant be afforded an impartial jury.”
Morgan v. Illinois,
The court gave each side ample opportunities to explore the venire members’ views — devoting five days (spanning nearly 900 pages of transcript) to the task. Nor did it restrict either side to abstract questions about whether a juror would follow instructions or perform his duties impartially,
cf. Morgan,
The trial court, it is true, drew the line at questions that sought to elicit the jurors’ views on Bedford’s specific case — but many judges understandably (and properly) would do the same thing to prevent the lawyers from previewing their case through voir dire.
Cf. United States v. Lawes,
These limitations did not render the process fundamentally unfair.
See Dennis,
B.
Bedford next'claims that the prosecutor’s closing arguments at the guilt and penalty phases violated due process. To prevail, Bedford must show that the prosecutor’s remarks were not just improper but that they were “flagrant.”
United States v. Carson,
Guilt phase. Bedford complains about comments by the prosecutor in closing argument at the guilt phase that allegedly disparaged defense counsel’s tactics. The prosecutor called “[s]ome” of Bed-ford’s “arguments” “Mickey Mouse defenses,” JA 2301, and he characterized others as attempts to “confuse” the jury by “flllfing] the courtroom with as much smoke as you possibly can,” JA 2304, casting aspersions “all around the courtroom” and putting “everyone on trial in the case except our little boy over here” — all in the hope that the jury would “lose sight of the real issues in the case,” JA 2315. Attempting to deflate an attempt by the defense to discredit a particular government witness, the prosecutor also predicted that the witness would “be dragged through the mud by the defense.” JA 2258.
These comments were not imрroper. The prosecution necessarily has “wide latitude” during closing argument to respond to the defense’s strategies, evidence and arguments.
United States v. Henry,
*234 Several of the prosecutor’s comments, Bedford adds, were calculated to incite the jury’s passions and were engineered to elicit an emotional, not a reasoned, reaction to the evidence. Responding to the defense theory that Bedford’s conduct was the culmination of an unplanned outburst — fueled by alcohol and emotion and sparked by a life-threatening confrontation with Toepfert’s new paramour — the prosecutor argued: (1) that the evidence, including graphic photographs of Toepfert’s and Smith’s bodies, proved Bed-ford’s conduct was purposeful and planned; (2) that Bedford’s inner “demon” — his alcohol dependence — was not responsible for his behavior, as' the only “demon in this case” was Bedford and (8) that the jurors’ duty required finding Bedford guilty and that, if they did so, each juror could say to himself “I did Gwen justice and I did Johnny justice,” JA 78.
These comments did not deprive Bed-ford of a fair trial. By alluding to the victim photographs, already admitted into evidence, and arguing that they established Bedford’s intent, the prosecutor permissibly sought to draw an inference from the evidence.
See Byrd,
Neither did the prosecutor overstep by urging the jury to do justice for Smith and Toepfert. Nothing prevents the government from appealing to the jurors’ sense of justice,
see Coe v. Bell,
Penalty phase.
In arguing that the prosecutor’s penalty-phase summation contained unfairly prejudicial comments, Bed-ford targets the following: (1) The prosecutor reminded the jury that they make only a recommendation, not a final decision, on Bedford’s sentence; (2) he read a passage from
Gregg v. Georgia,
We can quickly dispatch with Bed-ford’s complaints about the first four comments. There was nothing improper about accurately explaining to the jury that, under Ohio’s death-penalty scheme, they recommend — but do not definitively determine — the defendant’s sentence.
Hicks,
The prosecutor also did not overstep by using the pictures of the victims as evidence of an aggravating circumstance. True, only Bedford’s conviction for Toepfert’s aggravated murder carried a death specification,
see Bedford,
The remaining two statements require more -explanation. The prosecutor suggested that, even though under then-existing state law a life sentence would keep Bedford behind bars for 20 or 30 years beforе he could obtain parole, the law could change, enabling Bedford to obtain parole sooner. That statement is similar to informing the jury that if it selects a life sentence, state officials might commute the sentence to a shorter term. So long as the jury receives accurate information, it may consider the possibility, speculative though it may be, that future decisions of state executive officials could lead to the defendant’s early release.
See California v. Ramos,
Even if we assumed for the sake of argument that the prosecutor’s comments crossed the line, they were not sufficiently flagrant to make Bedford’s trial unfair.
Cf. Carson,
Bedford’s claim about the prosecutor’s comments on his unsworn statements fares no better. Under the Fifth
*236
(and Fourteenth) Amendment, the prosecution ordinarily may not comment on a defendant’s refusal to testify.
See Griffin v. California,
After noting that Bedford’s statement was unsworn and not subject to cross-examination, the prosecutor continued:
... I think because of that you can judge his credibility and the things that he had to say to you with a jaundice[d] eye because even if a person is under oath, you don’t have to believe what they say.... And the mere fact that this man elected to avoid being scrutinized by the prosecutor in this case should be considered by you.
JA 2434. Whether that comment was improper is a close call. The only point of allowing the prosecution to remind the jury that the defendant’s statement was not made under oath, after аll, is to enable the State (since it cannot cross-examine him) to challenge his credibility. At least the first part of the prosecutor’s comment seemed trained on that objective, encouraging the jury to question the truthfulness of what Bedford
did
say, not his refusal to testify under oath on another subject.
Cf. DePew,
Even assuming the prosecutor crossed the line, however, any violation was not flagrant. The likelihood that the jury was misled was low, as the court and the prosecutor told the jury Bedford was entitled to make an unsworn statement. The comment was isolated, and the prosecution’s other evidence was plentiful. The potential for prejudice was reduced still further by the fact that the state trial and appellate courts independently weighed the aggravating and mitigating circumstances.
Bedford,
Before turning to Bedford’s next argument, we must acknowledge one oddity about this analysis. Simply put, it is strangе to think of Bedford’s contention in conventional Fifth Amendment terms. The guarantee says that an individual shall not “be compelled in any criminal case to be a witness against himself.” U.S. Const, amend. V. Yet this issue arose not because the prosecution compelled Bedford to testify or because he exercised his right to remain silent and the prosecution disparaged his silence — the classic settings in which Fifth Amendment violations arise— but because Bedford
did
speak to the jury. Bedford voluntarily invoked a state allocu
*237
tion procedure that federal law does not require and that allowed him to make an unsworn statement to the jury during the penalty phase. At least ten States by our rough count have similar procedures that apply at the penalty phase of capital cases.
See Jeffries v. Blodgett,
Why any of this raises an issue of federal law in this case, however, is not self-evident. Doubtless, the invocation of the state-law right could implicate the Fifth Amendment if the prosecution insisted on cross-examining the defendant about
other
crimes.
See, e.g., DePew,
C.
Bedford next argues that the trial cоurt gave an unduly coercive Allen charge to the jury during the penalty phase. A day into its deliberations, the jury sent a note to the court asking “what would happen” if the jury could not reach a unanimous sentencing recommendation and whether there was “an approximate time frame” for reaching a decision. JA 2462. After consulting with the parties, the court responded with a supplemental instruction informing the jury that there was no fixed time limit but urging the jurors to “make every reasonable effort to agree on a recommendation,” given the time and energy already invested in the trial and the jurors’ superior position (having already participated in the guilt phase) to make a fair decision. JA 2468. The court “suggested]” that the jury first determine whether they were in fact deadlocked and, if so, to return a life-sentence recommendation. JA 2469.
The question is whether the instruction, viewed in context, was “coercive.”
Lowenfield v. Phelps,
By instructing the entire jury to make every reasonable effort to agree on a recommendation if they could do so in good faith, the trial court at least implicitly encouraged all of the jurors to reconsider their positions. Yes, the court did not explicitly direct the majority and the minority to do so. But that did not make the charge coercive. Reminding both sides of a split jury to remain open-minded no doubt may ensure that those in the minority are not singled out and pressured to acquiesce in the majority’s view,
see Williams v. Parke,
Nor was the charge coercive becаuse it omitted a caution that the jurors not abandon their honest convictions. There is no iron-clad rule that a trial court’s failure to include that reminder, though unfortunate and ill-advised, is invariably fatal to the conviction. See id. at 851. In this case, the trial court had instructed the jurors only the day before in its general charge not to “surrender honest convictions” in the interest of consensus. JA 2449. And although it could have done so more clearly, the court’s supplemental instruction alluded to the need to hold on to conscientiously held views. See JA 2468-69 (instructing the jurors “to make every reasonable conscientious] effort” to agree on a recommendation if possible).
The lack of other coercive language in the charge also diminished the need for an honest-conviction caveat. The reminder serves primarily to counterbalance the potentially coercive effect of the rest of the instruction, and the need for it depends on what sits on the other side of the scale. Here, the trial court responded to the jurors’ questions by informing them that there was no set time limit to reach a consensus, encouraging them to make reasonable efforts to agree and suggesting how they might proceed. The court never intimated that the jury had to reach agreement, instead explaining only what to do if consensus proved impossible.
Cf. Williams,
The trial court’s explanation of what would happen if the jury deadlocked also did not make the instruction coercive. The first part of the instruction, we realize, was inaccurate. It indicated that the court would declare a mistrial and call another jury in its place if the jury could not agree, even though Ohio law requires a judge confronted with an “irreconcilably deadlocked” jury to impose a life sentence, not empanel a new jury to start over.
See State v. Springer,
*239 D.
Bedford claims that his attorneys’ representation at both phases of the trial was constitutionally ineffective. To prevail, he must show that their performance was deficient and that, but for their poor performance, “there is a reasonable probability” the result would have been different.
Strickland v. Washington,
Bedford argues that his counsel should have objected to prosecutorial misconduct and incorrect jury instructions. As for their failure to object to the prosecutors’ guilt- and penalty-phase closing arguments, that did not constitute ineffective assistance because the comments were not flagrant.
See Slagle,
As for the jury instructions, even if the court’s directions were incorrect, they did not render his trial fundamentally unfair.
See Lawrence v. 48th Dist. Court,
Bedford also argues that the court should have given (and his counsel should have requested) an instruction at the outset that if the jury could not agree, they must impose a life sentence. But the trial court in fact instructed the jury to impose a life sentence if they could not agree. A death-penalty defendant, at any rate, has no constitutional entitlement to an instruction informing the jury of the effect of a deadlock.
See Jones v. United States,
Bedford also argues that his attorneys failed adequately to prepare and present mitigating evidence. In its detailed discussion of this claim, the district court explained why his counsel’s investigation was adequate and why their decisions about what witnesses and evidence to present reflected reasonable choices. We have nothing to add to its analysis on that score and cannot improve on it. We instead address only why the defense counsel’s conduct, even if it were deficient, did not prejudice Bedford.
To establish prejudice flowing from deficient penalty-phase preparation and presentation, the defendant must show that the evidence his attorneys should have discovered and put forward “differ[s] in a substantial way — in strength and subject matter — from the evidence actually presented.”
Hill v. Mitchell,
The jury heard mitigating evidencе from four witnesses. Dr. Donna Winter, a clinical psychologist who had examined Bed-ford, testified in the guilt phase that Bed- *240 ford had a history of chronic depression dating back a decade, exhibited “extreme stress” reflecting a “cry-for-help profile” in psychological tests, JA 2228, had a “borderline mentally retarded” IQ of 76, JA 2229, and would have had great difficulty handling the emotional and psychological strain (not helped by his alcohol consumption) caused by Toepfert’s rejection and his encounter with Smith. At the penalty phase, Dr. Nancy Schmidtgoessling elaborated on Bedford’s emotional instability— and the volatility his alcohol intake added to the mix — but she also underscored that he was “one of the more treatable” inmates she had encountered. JA 2352. Winter and Schmidtgoessling each prepared written reports — which they discussed in their testimony and which the jury was entitled to review, see Ohio Rev. Code § 2929.03(D)(1) — that fleshed out their findings and filled in details of Bed-ford’s troubled personal and family history.
Bedford’s attorneys also presented the testimony of Jackie Schmidt, a friend and former girlfriend of Bedford’s, who spent several hours with him the night before the murders and who confirmed his frayed emotional state. Bedford, who had been drinking a great deal, was deeply upset over Toepfert’s rejection, to the point of crying, and at one point he played Russian roulette with a loaded gun.
Finally, in his own (unsworn) testimony, Bedford told his life story, from his parents’ deaths while he was still young to his teenage marriage, his inability to care for his six children, his difficulty holding a job and maintaining healthy relationships and his long-running dependence on alcohol. Bedford described his relationship with Toepfert, his memory of the murders (much of which he could not recall) and his confession.
The evidence Bedford now says his lawyers overlooked (or never uncovered) does not differ “markedly” in strength or subject matter from the evidence they presented.
Hill,
Neither did the additional expert assessments differ substantially from what the jury heard. Schmidtgoessling, who was appointed to examine Bedford for purposes of a possible insanity plea, stated that, had she evaluated Bedford with mitigation in mind, she would have included additional facts from Bedford’s “social history.” JA 1694. But the facts she describes differ only marginally from the story the jury did hear, and although she might have connected the dots more closely, none of the links likely would have led the jury to a different conclusion.
Schmidtgoessling’s testimony, it is true, did not address the fact that Bedford allegedly had been “eating speed like candy” the two nights before the murders, JA 1691 — a fact Schmidtgoessling herself did not know when she testified. But Bed- *241 ford’s attorneys apparently were also unaware of this fact, as neither Jackie Schmidt (who brought this fact to light in her postconviction affidavit) nor Bedford himself told them about it when asked by counsel what drugs Bedford consumed before the murders. (Schmidt mentioned only alcohol, and Bedford mentioned alcohol and marijuana.)
Two other experts — Dr. Thomas Heiskell, a clinical psychologist, and Dr. James Tanley, a neuropsychologist — critiqued the accounts Schmidtgoessling and Winter offered at trial and suggested Bedford’s counsel should have arranged for more rigorous testing. But even assuming they are right, Bedford suffered no cognizable prejudice. Heiskell and Tanley raised only the
possibility
that more testing would have revealed other psychological or neurological problems, and neither highlighted any substantially different or stronger evidence of impairment that could have turned the tide. Each one, for instance, indicated that Bedford’s experts should have explored the possibility of оrganic brain damage, and Winter stated she would have pursued the possibility if she had received more information, but none of them concluded based on subsequent examinations that Bedford actually had brain damage at the time of the crime that more testing would have uncovered.
See Smith,
Bedford adds that the district court erred in denying his request for an evidentiary hearing regarding evidence his attorneys failed to explore or present.
See
28 U.S.C. § 2254(d) (1994). The issues he seeks to develop, however, pertain to the adequacy of his counsel’s performance, and there is no need to build a record on that issue because, as noted, he cannot show his attorneys’ performance prejudiced him.
See Ivory v. Jackson,
III.
For these reasons, we affirm.
