skip navigation

CONE v. BELL (No. 07-1114)
492 F. 3d 743, vacated and remanded.

HTML version
PDF version
HTML version
PDF version
HTML version
PDF version
HTML version
PDF version
HTML version
PDF version

Thomas, J., dissenting



on writ of certiorari to the united states court of appeals for the sixth circuit

[April 28, 2009]

    Justice Thomas, with whom Justice Scalia joins, dissenting.

    The Court affirms Gary Cone’s conviction for beating an elderly couple to death with a blunt object. In so doing, the majority correctly rejects Cone’s argument that his guilty verdict was secured in violation of his rights under Brady v. Maryland, 373 U. S. 83 (1963) . The majority declines, however, to decide whether the same evidence that was insufficient under Brady to overturn his conviction provides a basis for overturning his death sentence. The majority instead remands this question to the District Court for further consideration because it finds that the Court of Appeals engaged in a “summary treatment” of Cone’s Brady sentencing claim. See ante, at 25–27.

    I respectfully dissent. The Court of Appeals’ allegedly “summary treatment” of Cone’s sentencing claim does not justify a remand to the District Court. Cone has failed to establish “ ‘a reasonable probability that, had the evidence been disclosed to the defense, the result of the [sentencing] proceeding would have been different,’ ” Kyles v. Whitley, 514 U. S. 419, 435 (1995) (quoting United States v. Bagley, 473 U. S. 667, 682 (1985) (opinion of Blackmun, J.)). As a result, I would affirm the judgment of the Court of Appeals. 1


    This case arises from a crime spree 28 years ago that began with Cone’s robbery of a jewelry store in Memphis, Tennessee, and concluded with his robbery of a drugstore in Pompano Beach, Florida. Along the way, Cone shot a police officer and a bystander while trying to escape the first robbery, attempted to shoot another man in a failed carjacking attempt, unsuccessfully tried to force his way into a woman’s apartment at gunpoint, and murdered 93-year-old Shipley Todd and his 79-year-old wife, Cleopatra. When he was tried on two counts of first-degree murder in 1982, Cone’s sole defense was that he did not have the requisite intent to commit first-degree murder becausehe was in the grip of a chronic amphetamine psychosis. The jury rejected the defense and convicted Cone of both murders.

    At sentencing, the Tennessee jury found beyond a reasonable doubt that four statutory aggravating factors applied to Cone’s offense: (1) Cone had been convicted of one or more previous felonies involving the use or threat of violence; (2) he had knowingly created a great risk of death to two or more persons other than the victim during his act of murder; (3) the murder was especially heinous, atrocious or cruel in that it involved torture or depravity of mind; and (4) the murder was committed for the purpose of avoiding a lawful arrest. Tr. 2151–2152 (Apr. 23, 1982); see also State v. Cone, 665 S. W. 2d 87, 94–96 (Tenn. 1984). Tenn. Code Ann. §39–2-203(i) (1982).2 Cone argued to the jury at sentencing that his “capacity … to appreciate the wrongfulness of his conduct or to conform his conduct to the requirements of the law was substantially impaired as a result of mental disease or defect or intoxication which was insufficient to establish a defense to the crime but which substantially affected his judgment.” See §39–2-203(j)(8). But the jury found that neither this, nor any other mitigating factor, outweighed the aggravating factors. The jury, as required by Tennessee law, unanimously sentenced Cone to death. See §39–2-203(g).

    For almost three decades, Cone’s case has traveled through the Tennessee and federal courts. This Court has twice reversed decisions from the Court of Appeals that invalidated Cone’s conviction and sentence. See Bell v. Cone, 535 U. S. 685 (2002) ; Bell v. Cone, 543 U. S. 447 (2005) (per curiam). On remand from this Court’s latest decision, the Court of Appeals directly considered whether a handful of police reports, law enforcement bulletins, and notes that were allegedly withheld from Cone’s trial attorneys could have changed the result of Cone’s trial or sentencing. And, for the second time, the Court of Appeals held that there was not a “ ‘reasonable probability’ ” that the evidence would have altered the jury’s conclusion “that Cone’s prior drug use did not vitiate his specific intent to murder his victims and did not mitigate his culpability sufficient to avoid the death sentence.” 492 F. 3d 743, 757 (CA6 2007). The Court of Appeals, therefore, held that neither Cone’s conviction nor his sentence was invalid. See ibid.; Cone v. Bell, 243 F. 3d 961, 968 (CA6 2001). We should affirm the Court of Appeals and put an end to this litigation.


    According to the majority, the Court of Appeals’ decision affirming Cone’s death sentence is too “summary,” ante, at 25,and the facts are such that, on further examination, Cone “might” be able to demonstrate that it is “possible” that the contested evidence would have persuaded the jury to spare his life, ante, at 25–26. On this reasoning, the majority remands the case directly to the District Court for “full consideration [of] the merits of Cone’s [sentencing] claim.” Ante, at 27.I disagree on all counts. Remanding the sentencing issue to the District Court is an “unusual step” for this Court to take. House v. Bell, 547 U. S. 518, 557 (2006) (Roberts, C. J., concurring in judgment in part and dissenting in part). Furthermore, in this case, it is a step that is legally and factually unjustified. There is not “ ‘a reasonable probability that, had the evidence been disclosed to the defense, the result of the proceeding would have been different.’ ” Kyles, 514 U. S., at 433–434 (quoting Bagley, 473 U. S., at 682 (opinion of Blackmun, J.)).


    The majority’s criticism of the Court of Appeals’ allegedly “summary treatment” of the sentencing question is misplaced. Before the Court of Appeals, Cone dedicated eight pages of his opening brief to arguing that the implicated evidence was material to his guilt or innocence, but spent only one paragraph arguing its materiality to his death sentence. See Brief for Appellant in No. 99–5279 (CA6), pp. 40–48. The Court of Appeals’ focus on the guilt phase, rather than the sentencing phase, simply followed Cone’s lead. See 492 F. 3d, at 755 (“In his most recent brief, claiming that his receiving the withheld evidence would have resulted in a different sentence, Cone has made only conclusory arguments”).3 There is nothing defective about a judicial decision that summarily rejects an abbreviated legal argument, especially where, as here, the burden of proving the materiality of the contested evidence was on Cone.4


    In remanding this matter to the District Court, the majority makes two critical errors—one legal and one factual—that leave the false impression that Cone’s Brady claim has a chance of success. First, the majority states that “[i]t is possible that the suppressed evidence” may have convinced the jury that Cone’s substance abuse played a mitigating role in his crime and “[t]he evidence might also have rebutted the State’s suggestion” that Cone’s experts were inaccurately depicting the depth of his drug-induced impairment. Ante, at 26 (emphasis added); see also ante, at 26–27(remanding “[b]ecause the evidence suppressed at Cone’s trial may well have been material to the jury’s assessment of the proper punishment in this case” (emphasis added)). But, as the majority implicitly acknowledges, see ante, at 26, n. 19, this is not the correct legal test for evaluating a Brady claim: “The mere possibility that an item of undisclosed information might have helped the defense, or might have affected the outcome of the trial, does not establish ‘materiality’ in the constitutional sense.” United States v. Agurs, 427 U. S. 97, 109–110 (1976) (emphasis added).

    Rather,     this Court has made clear that the legal standard for adjudicating such a claim is whether there is a “reasonable probability” that the jury would have been persuaded by the allegedly withheld evidence. Kyles, supra, at 435; Bagley, supra, at 682 (opinion of Blackmun, J.). It simply is not sufficient, therefore, to claim that “there is a reasonable possibility that … testimony might have produced a different result … . [P]etitioner’s burden is to establish a reasonable probability of a different result.” Strickler v. Greene, 527 U. S. 263, 291 (1999) (emphasis in original). To satisfy the “reasonable probability” standard, Cone must show that “the favorable evidence could reasonably be taken to put the whole case in such a different light as to undermine confidence” in the jury’s sentencing determination. Kyles, supra, at 435. The Court must view the record “as a whole,” Sawyer v. Whitley, 505 U. S. 333, 374 (1992) (Stevens, J., concurring in judgment), and determine whether the absence of the disclosure prevented Cone from receiving “ ‘a trial resulting in a [sentence] worthy of confidence.’ ” Strickler, supra, at 290 (quoting Kyles, 514 U. S., at 434).

    In the context of this case, for Cone to establish “ ‘a reasonable probability that, had the evidence been disclosed to the defense, the result of the [sentencing] proceeding would have been different,’ ” id., at 435, he must not only demonstrate that the withheld evidence would have established that he was substantially impaired as a result of drug abuse or withdrawal; Cone also must establish that the addition of the allegedly withheld evidence ultimately would have led the jury to conclude that any mitigating factors (including substantial impairment) outweighed all of the established aggravating factors. See Tenn. Code Ann. §39–2-203(g).5

    Second, the majority incorrectly claims that to prevail on his Brady claim, Cone must demonstrate simply that the withheld evidence supported the inference that he “was impaired by his use of drugs around the time his crimes were committed.” See ante, at 21. This is factually inaccurate because there was already significant evidence of Cone’s drug use at trial. To establish that the allegedly withheld evidence would reasonably have had any impact on his case, Cone must instead show that the evidence would have supported his claim of substantial mental impairment from drug use.

    There was extensive evidence at trial that supported the inference that Cone was not only a longstanding drug user, but that he was in fact using drugs at the time of his crimes. The State itself presented significant evidence on this point. For example, it presented proof that officers found marijuana cigarette butts, empty drug vials, and loose syringes in the car that Cone abandoned immediately after the jewelry store robbery. Tr.1505–1509 (Apr. 19, 1982). The State also did not challenge testimony from Cone’s mother that Cone used drugs. Id., at 1647, 1648–1653 (Apr. 20, 1982). And, most tellingly, the State introduced evidence that Cone was abusing three drugs—cocaine, Dilaudid, and Demerol—at the time of his arrest and was suffering “slight withdrawal symptoms” from them. Id.,at 1915–1916, 1920 (Apr. 22, 1982). As the Court of Appeals explained, “[i]t would not have been news to the jurors, that Cone was a ‘drug user.’ ” 492 F. 3d, at 757.6

    In contrast, what was contested by the State during trial was Cone’s defense that his drug use was so significant that it caused him to suffer from extreme amphetamine psychosis at the time of the murders. One of Cone’s expert witnesses, a neuropharmacologist, testified that by the summer of 1980, when the crimes occurred, Cone was ingesting “ferociously large doses” of drugs and that his increasing tolerance and use of amphetamines caused a chronic amphetamine psychosis. Tr. 1736–1737, 1744–1747, 1758–1759 (Apr. 21, 1982). The expert further testified that if a person with chronic amphetamine psychosis were to go into withdrawal, he could suffer extreme mood swings, “a crashing depression,” and a state of weakness so severe that “he could barely lift himself.” Id., at 1857–1859. In this expert’s view, these symptoms could cause a person to “lose his mind.” Id., at 1859.

    The State contradicted that testimony with significant evidence that Cone did not act like someone who was “out of his mind” during the commission of his crimes. Rather, the State argued, Cone behaved rationally during his initial Tennessee robbery, his subsequent escape, his flight from Tennessee to Florida after the Todd murders, his Florida robbery, and his subsequent arrest. See, e.g., id., at 2074–2084 (Apr. 22, 1982). To substantiate this argument, the State called FBI Special Agent Eugene Flynn to the stand. Agent Flynn testified that, when captured, Cone coherently detailed his travel from Tennessee to Florida, explained his efforts to evade detection by shaving his beard and buying new clothes, and initiated negotiations for a plea bargain. Id., at 1918–1921. The State also presented testimony from a friend of Cone’s, Ilene Blankman, that she saw no indication that Cone was under the influence of drugs or severe withdrawal in the days immediately following the murder of the Todds. Id., at 1875–1876, 1882–1883 (Apr. 21, 1982).

    Viewing the record as a whole, then, it is apparent that the contested issue at trial and sentencing was not whether Cone used drugs, but rather the quantityof Cone’s drug use and its effect on his mental state. Only if the evidence allegedly withheld from Cone was relevant to this question whether Cone suffered from extreme amphetamine psychosis or other substantial impairment would the evidence have been exculpatory for purposes of Brady. See Order Denying Motion for Evidentiary Hearing and Order of Partial Dismissal, Cone v. Bell, No. 97–2312–M1/A (WD Tenn., May 15, 1998), App. to Pet. for Cert. 119a, n. 9 (explaining that “the issue at trial was not whether Cone had ever abused any drugs (he clearly had), but whether he was out of his mind on amphetamines at the time of the murders”); Tr. 2115–2116 (Apr. 23, 1982).


    With the legal and factual issues correctly framed, it becomes clear that Cone cannot establish a reasonable probability that admission of the evidence—viewed either individually or cumulatively—would have caused the jury to alter his sentence.



    Cone first argues that he was improperly denied police reports that included witness statements regarding Cone’s behavior around the time of his crime spree. The first statement was given by a convenience store employee, Robert McKinney, who saw Cone the day before he robbed the Tennessee jewelry store. When asked whether Cone appeared “to be drunk or high on anything,” McKinney answered, “[w]ell he did, he acted real weird … he just wandered around the store.” App. 49. But McKinney subsequently clarified that Cone “didn’t sound drunk” and that the reason Cone attracted his attention was because he “wasn’t acting like a regular customer”; he was “just kinda wandering” around the store. Motion to Expand the Record in No. 97–2312–M1 (WD Tenn.), Exh. 2, pp. 3, 4. Contrary to the majority’s assertion, this interview is not convincing evidence “that Cone appeared to be ‘drunk or high’ ” when McKinney saw him. Ante, at 21. McKinney’s clarification that he had characterized Cone’s behavior as “weird” because Cone appeared to be killing time rather than acting like a normal shopper undermines the implication of McKinney’s earlier statement that Cone looked “weird” because he might have been drunk or on drugs. Thus, there is little chance that McKinney’s statement would have provided any significant additional evidence that Cone was using drugs, let alone provide sentence-changing evidence that he was substantially impaired due to amphetamine psychosis.

    The second statement was given by Charles and Debbie Slaughter, who both witnessed Cone fleeing from police after the jewelry store robbery and reportedly told police that he looked “wild eyed.” App. 50. Cone had just robbed a jewelry store, shot a police officer and a bystander, and was still fleeing from police when seen by the Slaughters. It is thus unlikely that their observation of a “wild eyed” man would have been interpreted by the jury to mean that Cone “was suffering from chronic amphetamine psychosis at the time of the crimes,” ante, at 21, n. 16, rather than to mean that Cone looked like a man on the run.

    The third statement is contained in a police report authored by an officer who helped apprehend Cone after the Florida drugstore robbery. He reported that he saw a suspect “at the rear of Sambos restaurant. Subject was observed to be looking about in a frenzied manner and also appeared to be looking for a place to run.” App. 53. Nothing in this police report either connects Cone to drug use or appears otherwise capable of altering the jury’s understanding of Cone’s mental state at the time of the crimes. It certainly makes perfect sense that Cone was “looking about in a frenzied manner,” ibid.; he had just robbed a drugstore and was about to engage in a gun battle with police in order to evade arrest. The police officer’s description of Cone’s appearance under these circumstances thus does not “undermine confidence” in Cone’s sentence. Kyles, 514 U. S., at 435.


    The next category of documents that Cone relies upon to establish his Brady claim are police bulletins. Some of the bulletins were sent by Memphis Police Sergeant Roby to neighboring jurisdictions on the day of the Todd murders and the day after. The bulletins sought Cone’s apprehension and alternatively described him as a “drug user” or a “heavy drug user.” App. 55–58. Cone asserts that he could have used these bulletins to impeach Sergeant Roby’s trial testimony that the sergeant did not see any track marks when visiting Cone in jail a week later. Tr. 1939 (Apr. 22, 1982). Cone’s reasoning is faulty for two key reasons. First, Sergeant Roby never testified that Cone was nota drug user. His only trial testimony on this point was simply that he observed no “needle marks” on Cone’s arm when taking hair samples from him a few days after Cone’s apprehension. Ibid. Second, the bulletins establish only “that the police were initially cautious regarding the characteristics of a person who had committed several heinous crimes.” App. to Pet. for Cert. 119a, n. 9. The bulletins would not have tended to prove that the fugitive Cone was, in fact, a heavy drug user—let alone “out of his mind” or otherwise substantially impaired due to amphetamine psychosis—at the time of his crimes.7


    Cone also argues that material was withheld that could have been used to impeach Ilene Blankman’s testimony that Cone did not appear to be high or in withdrawal when she helped him obtain a Florida driver’s license during his efforts to evade arrest in Florida. Tr. 1875–1882 (Apr. 21, 1982). But he again fails to meet the standard for exculpatory evidence set by Brady.

    Cone first points to police notes of a pre-trial interview with Blankman, which did not reflect the statement she gave at trial that she saw no track marks on Cone’s arm. App. 72–73. But Blankman was questioned at trial about her failure to initially disclose this fact to police, Tr. 1903 (Apr. 21, 1982), so the jury was fully aware of the omission. Disclosure of the original copy of the police notes thus could not have had any material effect on the jury’s deliberations. Moreover, the missing notes also recorded a damning statement by Blankman that Cone “never used drugs around” her and she “never saw Cone with drug paraphernalia.” App. 73. Thus, it is difficult to accept Cone’s argument that he would have benefited from the introduction of notes from Blackman’s pretrial interview. If anything, these police notes would have undermined his mitigation argument.

    Cone next relies on a report that describes a woman’s confrontation with the prosecution team and Blankman at a restaurant during trial. During the encounter, the woman accused Blankman of lying on the stand in order to frame Cone for the murders. Id., at 74–75. The report indicates that the prosecutors politely declined the woman’s numerous attempts to discuss the merits of the case and that Blankman said nothing. Id., at 75. Nothing about this encounter raises doubts about Blankman’s credibility.

    Last, Cone points to “correspondence in the district attorney’s files suggest[ing] that the prosecution had been unusually solicitous of [Blankman’s] testimony.” Brief for Petitioner 45. But the correspondence was completely innocuous. One of the notes, sent in response to Blankman’s request for a copy of her prior statement, expressed to Blankman that her “cooperation in this particular matter is appreciated.” App. 76. The prosecutor then sent a letter to confirm that Blankman would testify at trial. Id., at 77. And finally, after trial, the prosecutor sent a note to inform Blankman of the verdict and indicate that they “certainly appreciate[d] [her] cooperation with [them] in the trial of Gary Bradford Cone.” Id., at 78. There is nothing about these notes that “tend[s] to prove any fact that is both favorable to Cone and material to his guilt or punishment.” App. to Pet. for Cert. 116a.


    Viewing the record as a whole, Cone has not come close to demonstrating that there is a “reasonable probability” that the withheld evidence, analyzed individually or cumulatively, would have changed the result of his sentencing. Much of the impeachment evidence identified by Cone is of no probative value whatsoever. The police bulletins do not contradict any of the trial testimony; the restaurant encounter was innocuous; and the correspondence sent by prosecutors to Blankman does not undermine her testimony or call Cone’s mental state into doubt. If the remaining evidence has any value to Cone, it is marginal at best. There was testimony that Blankman did not initially tell police that Cone lacked track marks. See Tr. 1903 (Apr. 21, 1982). McKinney clarified in his statement that Cone’s activity in the store was consistent with a person killing time, not the use of drugs or alcohol. And the behavior described by the Slaughters and the Florida police officer is more naturally attributable to the circumstances of Cone’s flight from the police than to any inference that Cone was “out of his mind” or otherwise substantially impaired due to amphetamine psychosis.

    Countering the trivial value of the alleged Brady material is the clear and overwhelming evidence that during Cone’s crime spree, he was neither sufficiently insane to avoid a conviction of murder nor substantially impaired by his drug use or withdrawal-related psychosis. There was substantial evidence that Cone carefully planned the jewelry store robbery and was calm in carrying it out, Tr. at 974–976, 1014 (Apr. 16, 1982), 1350–1352 (Apr. 17, 1982), 1501 (Apr. 19, 1982), 2075 (Apr. 22, 1982); that he successfully eluded police after engaging them in a shootout, id., at 1053–1064 (Apr. 16, 1982); that, after hiding overnight, he concocted a ruse to try to gain illegal entry to a residence, id., at 1205–1208 (Apr. 17, 1982); that he murdered the Todds after they declined to cooperate with his efforts to further elude police, id., at 1681 (Apr. 20, 1982); that he took steps to change his appearance at the Todd residence and then successfully fled to Florida, id., at 1918–1919 (Apr. 22, 1982); that he arrived in Florida exhibiting no signs of drug use or severe withdrawal, id., at 1875–1882 (Apr. 21, 1982); that he obtained false identification in a further effort to avoid apprehension, id., at 1881–1882, and that he denied any memory lapses and described undergoing only minor drug withdrawal when police arrested him, id., at 1919–1920 (Apr. 22, 1982). Given this wealth of evidence, there is no “reasonable probability” that the jury would have found that Cone was entitled to the substantial impairment mitigator had the evidence he seeks been made available to him.

    And even if Cone could have presented this evidence to the jury at sentencing and established an entitlement to this mitigator, he still has not demonstrated a reasonable probability that it would have outweighed all of the aggravating factors supporting the jury’s death sentence. See id., at 2151–2154 (Apr. 23, 1982). In its decision on direct appeal, the Tennessee Supreme Court was well aware of the evidence regarding the “degree and extent of [Cone’s] drug abuse.” Cone, 665 S. W. 2d, at 90. As part of its required independent review of whether the mitigation evidence was sufficiently substantial to outweigh the aggravating factors, see Tenn. Code Ann. §39–2-205, the Tennessee court nevertheless concluded that the sentence was “not in any way disproportionate under all of the circumstances, including the brutal murders of two elderly defenseless persons by an escaping armed robber who had terrorized a residential neighborhood for twenty-four hours.” 665 S. W. 2d, at 95–96. None of Cone’s proffered evidence places that conclusion, made by both the jury and the Tennessee Supreme Court, “in such a different light as to undermine confidence” in Cone’s sentence. Kyles, 514 U. S., at 435; see also Strickler, 527 U. S., at 296.


    This Court should not vacate and remand lower court decisions based on nothing more than the vague suspicion that error might be present, or because the court below could have been more clear. This is especially so where, as here, the record before the Court is adequate to evaluate Cone’s Brady claims with respect to both the guilt and sentencing phases of his trial. The Court’s willingness to return the sentencing issue to the District Court without any firm conviction that an error was committed by the Court of Appeals is inconsistent with our established practice and disrespectful to the lower courts that have considered this case. Worse still, the inevitable result will be years of additional delay in the execution of a death sentence lawfully imposed by a Tennessee jury. Because I would affirm the judgment below, I respectfully dissent.


1 Because I would affirm on the basis of the Court of Appeals’ alternative holding below, I do not reach the issues of procedural default resolved by the majority. See United States v. Atlantic Research Corp., 551 U. S. 128, 141, n. 8 (2007) ; Ayotte v. Planned Parenthood of Northern New Eng., 546 U. S. 320, 332 (2006) ; Ardestani v. INS, 502 U. S. 129, 139 (1991) .

2 The Tennessee Supreme Court later concluded that the record in Cone’s case was doubtful as to evidence supporting the second circumstance given the lapse in time between the initial events of the escape and the Todd murders. Cone, 665 S. W. 2d, at 95. The court, however, determined that the existence of the other three factors rendered any possible error in this factor harmless beyond a reasonable doubt. Ibid.

3 The assertion by the majority, ante, at 26, n. 19, and Justice Alito, ante, at 8 (opinion concurring in part and dissenting in part), that the Court of Appeals did not address the merits of the sentencing issue at all is flatly wrong. See 492 F. 3d, at 757 (rejecting Cone’s Brady claim because the proffered evidence would not have altered the jury’s conclusion “that Cone’s prior drug use did not vitiate his specific intent to murder his victims and did not mitigate his culpability sufficient to avoid the death sentence” (emphasis added)).

4 The majority does not attempt to justify its remand by contending that it is necessary because the record is insufficient to decide the claim. Nor could it persuasively contend a remand is necessary so that the District Court can hold an evidentiary hearing. Such a hearing would shed no additional light on the trial proceedings or the relative impeachment value of the withheld documents. Cone himself agrees that “this Court should resolve the merits of [his] Brady claim.” Reply Brief for Petitioner 24; see also Brief for Respondent 26–27.

5 The majority asserts that the standard under Tennessee law for demonstrating mental defect or intoxication as a mitigating factor at sentencing is “far lesser” than the standard for demonstrating insanity in the guilt phase of a criminal trial. Ante, at 25. But the mitigating factor still requires a showing that Cone’s mental capacity was “substantially impaired” as a result of mental defect. Tenn. Code Ann. §39–2-203(j)(8). In any event, the only authority cited by the majority for its assertion that the standard is “far” lesser than that for insanity is Justice Stevens’ lone dissent in a prior appeal in this case. Ante, at 25.

6 Although there were two occasions during closing arguments where prosecutors intimated that Cone was not a drug user, see Tr. 2014–2015, 2068 (Apr. 22, 1982), the State’s argument otherwise consistently focused on the real issue in the case: that Cone was not so significantly affected by his drug use around the time of his crimes that he was “out of his mind” or “drug crazy” during the critical days of August 1980. See id., at 2023–2024, 2071–2084. The majority’s focus on two brief excerpts from the State’s closing argument fails to faithfully view the record “as a whole” for purposes of a Brady analysis. See Sawyer v. Whitley, 505 U. S. 333, 374 (1992) (Stevens, J., concurring in judgment); see also Strickler v. Greene, 527 U. S. 263, 290–291 (1999) (finding no reasonable probability of a different result even when prosecutor’s closing argument relied on testimony that could have been impeached by withheld material).

7 Alert bulletins sent by the FBI similarly identified Cone as a “believed heavy drug user” or a “drug user.” App. 62–70. Cone argues that these bulletins could have been used to impeach FBI Agent Flynn’s testimony about Cone’s arrest in Florida. The bulletins would not have constituted material impeachment evidence, however, for the second reason identified above. In addition, the bulletins would not have contradicted any of FBI Agent Flynn’s testimony; he in fact stated at trial that Cone reported using three drugs and was undergoing mild drug withdrawal when he was captured in Florida. Tr. 1915–1916 (Apr. 22, 1982).