On March 3, 1974, after a jury trial in the Superior Court of Colquitt County, Georgia, appellant Willie X. Ross was convicted of armed robbery, kidnapping and murder. Sentences of life imprisonment, twenty years and the death penalty were imposed, respectively. The convictions and sentences were affirmed by the Georgia Supreme Court. Ross v. State, 233 Ga. 361, 211 S.E.2d 356 (1974), cert. denied, 428 U.S. 910, 96 S. Ct. 3222, 49 L. Ed. 2d 1217 (1976). Ross petitioned for state habeas corpus relief, but, after a hearing, relief was denied. That decision was affirmed in Ross v. Hopper, 240 Ga. 369, 240 S.E.2d 850 (1977), cert. denied, 435 U.S. 1018, 98 S. Ct. 1890, 56 L. Ed. 2d 397 (1978). Meanwhile, Ross filed a petition for declaratory judgment in the superior court, alleging that his trial jury was influenced by doubt as to the constitutional validity of the Georgia death penalty statute. The petition for declaratory judgment was denied, and the Georgia Supreme Court affirmed. Ross v. State, 238 Ga. 445, 233 S.E.2d 381 (1977).

Ross then filed a petition for federal habeas corpus relief in the United States District Court for the Southern District of Georgia. The matter was consolidated with two other cases in which the death penalty had been imposed, Mitchell v. Hopper, CV No. 478-132, and Spencer v. Zant, CV No. 179-247. The three petitions were denied. Mitchell v. Hopper, 538 F. Supp. 77 (S.D.Ga.1982); Ross v. Hopper, 538 F. Supp. 105 (S.D.Ga.1982). On appeal, a panel of this court affirmed in part the denial of Ross' petition, but reversed and remanded for further evidentiary development on the issue of whether the Georgia death penalty statute was being applied arbitrarily and discriminatorily to blacks. Ross v. Hopper, 716 F.2d 1528 (11th Cir.1983). A majority of this court subsequently voted to rehear Ross' appeal en banc, and the panel opinion was vacated. Ross v. Hopper, 729 F.2d 1293 (11th Cir.1984). On rehearing en banc, we again affirm in part the district court's denial of Ross' petition, but we remand to the panel for further consideration of Ross' claim relating to the jury composition, in light of Ross' motion to supplement the record.

On August 23, 1973, appellant Willie X. Ross, Freddie Lee King, Rudy Turner, and Theodore Ross, appellant's brother, drove from Madison, Florida to the Clover Farms Highway Grocery at Moultrie, Georgia. When the store closed for the evening they followed the individual who closed the store to a nearby house in which the J.R. Stanford family lived. They then drove back to Madison.

The next evening, August 24, the four men returned to the Stanford home. Wearing stocking masks over their faces, they entered the home, held the family at gunpoint and went through the house collecting various valuables, including Mr. Stanford's .32 caliber pistol. Upon demanding the money from the grocery store, they were told it was in the possession of Robert Lee, who lived nearby, and that Wendell Norman, Stanford's son-in-law and Lee's partner in the grocery store, would return to the Stanford home later that night. When Norman arrived, he was ordered by the intruders to take Theodore Ross and King to get the money. Stanford's fourteen-year-old stepdaughter was taken also as hostage. Appellant Willie Ross and Turner remained at the Stanford home.

Theodore Ross and King, with Norman and the stepdaughter, drove to Lee's home, entered and proceeded to Lee's bedroom. When Norman awoke Lee and explained why they were there, Lee reached for his pistol and fired into the hallway. Either Theodore Ross or King returned fire and grabbed one of Lee's small sons, threatening to kill the child if Lee did not stop firing and turn over the money. Theodore Ross and King were given the cash box containing approximately $20,000 in cash and checks, and fled on foot. Norman then contacted the police.

Lieutenant Tommie Meredith of the Moultrie Police Department responded to the call and drove to the Stanford home, closely followed in a separate car by another officer. Members of the Stanford family testified that Meredith, armed with a shotgun, entered through the kitchen door confronting Turner, who, armed with a .22 caliber pistol, was crouching at the opposite end of a table in the adjoining dining room. Ross was seen standing against a wall near a refrigerator in the dining room and armed with Stanford's .32 caliber pistol. Turner, stating, "I've got them right here," motioned for one of the family members to come toward him. The Stanfords, however, fled to a bedroom and closed the door. Immediately thereafter, both Mr. Stanford and the other police officer heard an exchange of gunfire. The officer, approaching the house from outside, saw and fired at two persons running from the house through the back yard.

The officer found Lieutenant Meredith's body on the kitchen floor, shot through the chest at point-blank range. The pistol last seen in Turner's possession was discovered in the back yard, fully loaded, the cartridge in the firing chamber bearing an indentation indicating the pistol had misfired. The .32 caliber pistol belonging to Stanford and last seen in Willie Ross' hand seconds before the shooting, was found near the back-yard fence, one round having been fired from it. A microanalyst for the State Crime Laboratory identified the bullet removed from Meredith's body as having been fired from the .32 caliber pistol.

Theodore Ross testified at appellant's trial that appellant had told him that he (Willie Ross) thought he had shot a policeman and that Turner's gun had misfired. Bobby Gamble, another prosecution witness, who had driven appellant back to Florida in the days subsequent to the incident, also testified that appellant had told him he thought he had killed a policeman.

Appellant was apprehended in New York several months after the incident, extradited to Georgia, and indicted for kidnapping, armed robbery, and murder. Appellant was convicted on all counts, and was sentenced to death for the murder of Lieutenant Meredith.II. The Claims

Appellant challenges the district court's resolution of six of the issues raised in his federal habeas corpus petition. Appellant argues that (1) under the authority of Enmund v. Florida, 458 U.S. 782, 102 S. Ct. 3368, 73 L. Ed. 2d 1140 (1982), the death penalty constitutionally may not be imposed when appellant was convicted of felony murder and there was no specific finding by the jury that he killed, attempted to kill, or intended to kill; (2) the precepts of Brady v. Maryland, 373 U.S. 83, 83 S. Ct. 1194, 10 L. Ed. 2d 215 (1963), were violated by the state's knowing use of perjured testimony of appellant's brother Theodore that appellant had told Theodore he had shot a policeman; (3) the precepts of Giglio v. United States, 405 U.S. 150, 92 S. Ct. 763, 31 L. Ed. 2d 104 (1972), were violated by the state's failure to disclose evidence contradicting Theodore's statement that no promises had been made to him in exchange for his testimony against appellant; (4) the district court erroneously failed to hold an evidentiary hearing on appellant's challenge to the composition of the grand and traverse juries;2 (5) an evidentiary hearing also was required in regard to appellant's claim that the Georgia death penalty statute is applied in an arbitrary and racially discriminatory manner and with inadequate appellate review; and (6) the district court should have conducted an evidentiary hearing on appellant's claim that failure to grant a change of venue due to prejudicial pretrial publicity deprived him of a fair trial. In addition, appellant contends for the first time on appeal that the charge to the jury in the sentencing phase of the trial did not adequately explain the function of mitigating circumstances, in violation of Spivey v. Zant, 661 F.2d 464 (5th Cir.1981), cert. denied, 458 U.S. 1111, 102 S. Ct. 3495, 73 L. Ed. 2d 1374 (1982); Goodwin v. Balkcom, 684 F.2d 794 (11th Cir.1982), cert. denied, 460 U.S. 1098, 103 S. Ct. 1798, 76 L. Ed. 2d 364 (1983); and Westbrook v. Zant, 704 F.2d 1487 (11th Cir.1983).

On rehearing en banc, we agree with the panel's disposition of appellant's Brady, Giglio, and change of venue claims, along with appellant's challenge to the jury charge in the sentencing phase of the trial. We therefore reinstate sections II.B., II.C., II.F., and II.G. of the panel opinion. See Ross v. Hopper, 716 F.2d at 1533-37, 1539-42. We proceed to discuss appellant's remaining claims.

Appellant's first remaining claim is based on Enmund v. Florida, 458 U.S. 782, 102 S. Ct. 3368, 73 L. Ed. 2d 1140 (1982), where the Supreme Court held that the eighth amendment barred the imposition of the death penalty "in the absence of proof that [the defendant] killed or attempted to kill, and regardless of whether [he] intended or contemplated that life be taken...." Id. AT 801, 102 S.Ct. at 3379. Appellant contends that Enmund 's holding was violated in his case because the jury did not make a specific finding that he either murdered or intended to murder Lieutenant Meredith, and, therefore, he is entitled to a new sentencing hearing. Because we do not read Enmund as constitutionally requiring that a jury expressly make such a culpability determination, and because we find that the death penalty is proportional to appellant's conduct under Enmund, we affirm the district court's denial of habeas relief.

In Enmund, the defendant was convicted and sentenced to death for murders committed by his co-felons during a robbery. The Florida Supreme Court affirmed the conviction and sentence on the theory that the jury could have found that Enmund drove the getaway car and was thus a principal responsible for his co-felons' acts. Enmund v. State, 399 So. 2d 1362, 1369-70 (Fla.1981). The Florida Supreme Court further concluded that, although the evidence at most supported an inference that Enmund had driven the getaway car, there was no constitutional bar to imposing the death penalty even though the evidence did not show that the defendant intended that life be taken. Id. at 1371.

The United States Supreme Court reversed the defendant's sentence, holding that the eighth amendment does not permit the imposition of the death penalty "on one such as Enmund who aids and abets a felony in the course of which a murder is committed by others but who does not himself kill, attempt to kill, or intend that a killing take place or that lethal force will be employed." Enmund, 102 S. Ct. at 3376-77. The Court's holding was based on the principle that a defendant's sentence must not be grossly disproportionate to his crime. Id. at 3376 ("Petitioner's argument is that because he did not kill, attempt to kill, and he did not intend to kill, the death penalty is disproportionate as applied to him...." (emphasis in original) ).3 The Court also reasoned that the two principal social purposes of the death penalty--deterrence and retribution--would not be furthered by imposing the death penalty on a defendant who did not intend or contemplate that murder would take place. Deterrence would not be furthered because, quite simply, a person cannot be deterred who does not intend or contemplate that the acts being punished would occur. Id. at 3377-78. Similarly, retribution would not be justified where the punishment did not correspond with the defendant's culpability--"what [his] intentions, expectations, and actions were," id. at 3378, and where the defendant did not intend or cause the killings, the retributive purpose would not be served by executing the defendant. Therefore, because the record as construed by the Florida Supreme Court did "not warrant a finding that Enmund had any intention of participating in or facilitating a murder," id. at 3377, the United States Supreme Court held that his sentence of death violated the eighth amendment.

Although the Court in Enmund outlined a minimum threshold beneath which the defendant's culpability must not fall if the death penalty is to be imposed, it did not delineate how or by whom the determination was to be made. Relying on cases holding that the question of the defendant's guilt is solely for the jury, see, e.g., Bollenbach v. United States, 326 U.S. 607, 614, 66 S. Ct. 402, 614, 90 L. Ed. 350 (1946), appellant contends that the determination of whether the defendant's culpability satisfies Enmund must be made by the jury and not by an appellate court from the record. Appellant notes that at least one other circuit has required that a new sentencing hearing be held where it is unclear from the jury's verdict whether the jury found Enmund 's threshold of culpability satisfied. See Bullock v. Lucas, 743 F.2d 244 (5th Cir.1984), petition for cert. filed, 53 U.S.L.W. 3568 (U.S. Jan. 29, 1985) (No. 84-1236); Jones v. Thigpen, 741 F.2d 805 (5th Cir.1984), petition for cert. filed, 53 U.S.L.W. 3568 (U.S. Jan. 30, 1985) (No. 84-1237); Reddix v. Thigpen, 728 F.2d 705 (5th Cir.), reh'g denied, 732 F.2d 494 (5th Cir.1984), cert. denied, --- U.S. ----, 105 S. Ct. 397, 83 L. Ed. 2d 331 (1984); see also Garcia v. Illinois, --- U.S. ----, 104 S. Ct. 3555, 82 L. Ed. 2d 856 (1984) (Marshall, J., dissenting from denial of certiorari). But see State v. Tison, 142 Ariz. 446, 690 P.2d 747, 748-49 (1984) (en banc); People v. Garcia, 97 Ill. 2d 58, 73 Ill. Dec. 414, 454 N.E.2d 274, 284-85 (1983), cert. denied, --- U.S. ----, 104 S. Ct. 3555, 82 L. Ed. 2d 856 (1984).

In our opinion, the question of whether the defendant's culpability satisfies the eighth amendment is sufficiently distinct from the question of the defendant's guilt that a specific jury finding is not constitutionally required. The necessity of ensuring that the jury finds every element of the crime beyond a reasonable doubt focuses on procedural due process concerns, see United States v. U.S. Gypsum Co., 438 U.S. 422, 98 S. Ct. 2864, 57 L. Ed. 2d 854 (1978), whereas the proportionality review mandated by Enmund is aimed at ensuring that imposition of the death penalty is not excessive in light of the defendant's culpability. 102 S. Ct. at 3377. We see no reason why the Enmund inquiry need be characterized as an additional element that the state must prove to the jury beyond a reasonable doubt at the guilt or sentencing phase before the death penalty lawfully can be imposed, so long as a reviewing court can determine from the record that the eighth amendment has been satisfied. It may very well be that a specific jury finding on the defendant's intent or an "Enmund instruction" at the sentencing hearing is desirable, to develop a record that will assist a reviewing court in conducting a proper proportionality review in a particular case. We decline to transform the desirability of such procedures, however, into a constitutional requirement that the trier of fact make specific Enmund findings, such that every defendant sentenced to death without express jury findings on culpability is entitled to a new sentencing hearing.4

Thus, we conclude that specific jury findings on the defendant's culpability are not a necessary constitutional corollary to the Court's holding in Enmund. Consequently, appellant is not entitled to a new sentencing hearing in this case solely because his jury returned a general guilty verdict after being instructed on felony murder and conspiracy. Rather, just as the Supreme Court in Enmund reviewed the record as construed by the Florida Supreme Court, we also must turn to the record to determine whether appellant's sentence violates the eighth amendment.5

In Enmund, the Supreme Court reversed the defendant's death penalty because it had been imposed "in the absence of proof that Enmund killed or attempted to kill, and regardless of whether Enmund intended or contemplated that life be taken...." 102 S. Ct. at 3379 (emphasis supplied). Appellant's involvement in this case stands in stark contrast to Enmund's role as the driver of a getaway car. Even taking all of the evidence most favorably to appellant, we conclude that his culpability as shown by the record satisfies the eighth amendment standard set out in Enmund.

Although the jury was charged on felony murder, the state's theory at trial was that appellant was the triggerman. Testimony showed that, when Lieutenant Meredith entered the house, appellant was positioned in the dining room and armed with a .32 caliber pistol stolen from Stanford. Turner was also armed with a pistol, crouched at the end of the dining room table away from the kitchen. Witnesses heard two shots from Meredith's shotgun, followed by a single pistol shot. Turner and appellant fled through the backyard while being fired upon by Officer Lynch, who had arrived shortly after Lieutenant Meredith. Two guns were later found in the backyard: Turner's pistol, which was fully loaded and had a cartridge in the firing chamber that was indented, indicating it had misfired, and the .32 caliber pistol, last seen in appellant's possession, which ballistic tests later showed had fired the shot that killed Meredith. At trial, appellant's brother testified that appellant had told him that he had shot a policeman and that Turner's gun had misfired. Another witness also testified that appellant had stated that he thought he had killed a policeman.

The above summary of the evidence presented at trial, as recounted by the Georgia Supreme Court, Ross v. State, 233 Ga. 361, 364, 211 S.E.2d 356, 358 (1974), would fully warrant a conclusion that appellant fired the shot that killed Lieutenant Meredith. Appellant, however, suggests that the possibility exists that Turner, upon having his pistol misfire, may have grabbed the .32 caliber pistol and fired the fatal shot instead of appellant. Appellant also relies on his brother's later recantation of his testimony that appellant had told him that he had shot a policeman. Even accepting appellant's rather strained version of events and the remote possibility that appellant did not himself fire the fatal shot, we believe that the uncontroverted evidence in the record shows that appellant's sentence of death does not violate the eighth amendment.

We do not read Enmund as barring the death penalty for all non-triggermen, but merely as requiring a level of individual participation that justifies the application of the death penalty. 102 S. Ct. at 3377. The Supreme Court's objection in Enmund that Earl Enmund was to be executed "regardless of whether [he] intended or contemplated that life would be taken," id. at 3379, simply does not extend to appellant, whose actions undeniably reflect the contemplation that life would be taken. Appellant does not deny that he and Turner held the Stanford family hostage while their cohorts went to the Lee home. Likewise, it is undisputed that appellant was armed and in the dining room with Turner when Lieutenant Meredith entered the home. He was thus actively engaged in furthering the course of events that led directly to Meredith's murder, whether or not he actually pulled the trigger.

More fundamentally, it is evident that the two primary purposes of capital punishment--deterrence and retribution--legitimately can be applied to the facts of this case. The Enmund Court observed that the death penalty will not likely deter "if a person does not intend that life be taken or contemplate that lethal force will be employed by others...." Id. at 3377. Here, however, it would be incredible to believe that appellant did not contemplate that lethal force would be used either by himself or his accomplice. Likewise, the "intentions, expectations, and actions," id. at 3378, of an individual engaging in such acts rise to a level of culpability such that the retributive purposes of capital punishment are furthered by appellant's sentence of death.

In sum, even if appellant's actions are viewed in their most favorable light, and even if we completely discount the testimony of those witnesses who stated that appellant had told them he thought he had shot a policeman, appellant's culpability still is of a magnitude wholly different from that of Earl Enmund. In Enmund, the evidence showed at most that the defendant had been seated in a getaway car at the side of the road awaiting the robbers' escape. In contrast, the record here depicts an individual who undoubtedly contemplated that lethal force would be used either by himself or by others as they held a family hostage, and who actively participated in the activities that culminated in Lieutenant Meredith's death. Consequently, we find that the death penalty in this case does not violate the eighth amendment, and we affirm the district court's denial of habeas corpus relief on this ground.

Appellant's next claim is that the venires from which the grand and traverse juries that indicted, convicted and sentenced him were selected, substantially underrepresented blacks and women. The district court rejected this claim because appellant "was afforded a full and fair opportunity to present his jury composition issue to the courts of Georgia and ... simply failed to meet his burden of proof." Ross v. Hopper, 538 F. Supp. at 106. The district court therefore applied the presumption of correctness created by 28 U.S.C. Sec. 2254(d) to the state courts' determination that appellant's claim had no factual basis. Id. The panel affirmed, noting that appellant did not establish that his claim fell within one of the exceptions to the statutory presumption of correctness listed in Townsend v. Sain, 372 U.S. 293, 83 S. Ct. 745, 9 L. Ed. 2d 770 (1963). See Ross v. Hopper, 716 F.2d at 1537-39.

On the same day this court reheard the instant case en banc, appellant filed a motion to supplement the record concerning his jury composition claim. Because the evidence now proffered by appellant may affect the applicability of one or more of the Townsend exceptions, we remand to the panel for consideration of this claim in light of appellant's motion to supplement the record.

C. The Arbitrary and Racially Discriminatory Application Claim

Appellant's final claim is that the Georgia death penalty statute is applied in an arbitrary and racially discriminatory manner and with inadequate appellate review. Appellant first raised this claim in his state habeas corpus petition in 1976.6 At an evidentiary hearing on December 9, 1976, appellant introduced testimony by Dr. Tobe Johnson, a professor of political science at Morehouse College. Dr. Johnson testified concerning his review of two social science studies on racial discrimination in the application of the death penalty in Georgia.7 On March 22, 1977, the superior court denied appellant's habeas petition. Appellant moved to reopen the evidentiary hearing in order to introduce a 1972 study of capital punishment in Georgia conducted by the Georgia Department of Corrections. The motion was denied.

The Georgia Supreme Court affirmed the superior court's denial of habeas relief. Ross v. Hopper, 240 Ga. 369, 240 S.E.2d 850 (1977), cert. denied, 435 U.S. 1018, 98 S. Ct. 1890, 56 L. Ed. 2d 397 (1978). The court noted that "[Dr. Johnson's] testimony was largely based on his analysis of data gathered prior to the decision in Furman v. Georgia, 408 U.S. 238 [92 S. Ct. 2726, 33 L. Ed. 2d 346] (1972)," and that the Georgia Department of Corrections study "was based on a study of capital punishment in Georgia for the years 1943 to 1965" and therefore was "irrelevant to this case since [Ross] was convicted and sentenced after Furman v. Georgia." Ross, 240 S.E.2d at 852-53. The court concluded that appellant's evidence "was insufficient to show that the death penalty was arbitrarily and capriciously imposed in his case as a result of deliberate discrimination...." Id. at 853.

Appellant also asserted the claim in his federal habeas corpus petition in 1978. Appellant's petition was consolidated with those of two other petitioners, James Lee Spencer and William "Billy" Mitchell, both of whom raised the same claim. At a hearing on January 26, 1981, counsel for all three petitioners proffered a study, conducted by Dr. William Bowers and Glenn Pierce, of criminal homicides and the death penalty in Georgia and other states between July, 1972, and August, 1980 (the "Bowers and Pierce study").8 Counsel also proffered a study, conducted by University of Texas law professor George Dix, of appellate review in cases where the death penalty was imposed (the "Dix study").9 The district court denied counsel's request for an evidentiary hearing to introduce the proffered studies.

On April 6, 1982, the court entered final judgment denying the three consolidated petitions. Mitchell v. Hopper, 538 F. Supp. 77 (S.D.Ga.1982); Ross v. Hopper, 538 F. Supp. 105 (S.D.Ga.1982). On April 15, appellant filed a memorandum of law in support of a motion to alter or amend the judgment of the court. In the memorandum of law, appellant proffered a study, conducted by Dr. David C. Baldus, of the application of the death penalty in Georgia (the "Baldus study").10 The motion to alter or amend the judgment was denied.

On appeal from the denial of Spencer's petition, a panel of this court held, inter alia, that the district court erred in refusing counsel's request for an evidentiary hearing to introduce the Baldus study. Spencer v. Zant, 715 F.2d 1562, 1578-83 (11th Cir.), en banc reh'g granted, 715 F.2d 1583 (11th Cir.1983), decision withheld pending submission and en banc consideration of McCleskey v. Zant and Ross v. Hopper, 729 F.2d 1293, 1294 (11th Cir.1984). The panel therefore vacated the judgment and remanded for an evidentiary hearing. Spencer v. Zant, 715 F.2d at 1583. On appeal from the denial of Ross' petition, the same panel, relying on Spencer v. Zant, reversed and remanded "for further evidentiary development" as to the arbitrary and racially discriminatory application claim. Ross v. Hopper, 716 F.2d 1528, 1539 (11th Cir.1983), en banc reh'g granted, 729 F.2d 1293, 1294 (11th Cir.1984).

Upon reevaluation of appellant's claim in light of our recent en banc opinion in McCleskey v. Kemp, 753 F.2d 877 (11th Cir.1985), we hold that the district court did not err when it denied appellant's request for an evidentiary hearing. In McCleskey, we explained that statistical evidence of a disparate impact in the imposition of the death penalty is relevant only to the extent that it compels an inference of purposeful discrimination:

[P]roof of a disparate impact alone is insufficient to invalidate a capital sentencing system, unless that disparate impact is so great that it compels a conclusion that the system is unprincipled, irrational, arbitrary and capricious such that purposeful discrimination--i.e., race is intentionally being used as a factor in sentencing--can be presumed to permeate the system.

Id. at 892.

Hence, when a petitioner proffers statistical evidence of the kind relied upon by appellant in the instant case, the district court need not hold an evidentiary hearing unless the proffer demonstrates a reasonable possibility that the evidence might compel an inference of purposeful discrimination. As we stated in McCleskey:

Needless to say, an evidentiary hearing would be necessary to hear any evidence that a particular defendant was discriminated against because of his race. But general statistical studies of the kind offered here do not even purport to prove that fact. Aside from that kind of evidence, however, it would not seem necessary to conduct a full evidentiary hearing as to studies which do nothing more than show an unexplainable disparity. Generalized studies would appear to have little hope of excluding every possible factor that might make a difference between crimes and defendants, exclusive of race. To the extent there is a subjective or judgmental component to the discretion with which a sentence is invested, not only will no two defendants be seen identical by the sentencers, but no two sentencers will see a single case precisely the same. As the court has recognized, there are "countless racially neutral variables" in the sentencing of capital cases. Smith v. Balkcom, 671 F.2d [858,] 859 [ (5th Cir. Unit B), cert. denied, 459 U.S. 882, 103 S. Ct. 181, 74 L. Ed. 2d 148 (1982) ].

The proffer in the instant case falls far short of demonstrating a reasonable possibility that the evidence might compel such an inference. With respect to the studies raised in the state habeas corpus proceedings, we agree with the Georgia Supreme Court that, because those studies were based on data gathered prior to Furman v. Georgia, they do not support appellant's claim. Statistical evidence of disparate impact in the pre-Furman Georgia capital sentencing system can hardly compel an inference of purposeful discrimination in the post-Furman system.

With respect to the Baldus study, McCleskey is squarely controlling. Because the district court in McCleskey held an extensive evidentiary hearing on the Baldus study, this court had the opportunity to consider, on the basis of a fully developed record, both the methodology of the study and the inferences to be drawn from it. We discussed the strengths and weaknesses of statistical evidence in general, and the Baldus study in particular, before concluding that "the Baldus study is simply insufficient to support a ruling, in the context of a statute that is operating much as intended, that racial factors are playing a role in the outcome sufficient to render the system as a whole arbitrary and capricious." McCleskey, 753 F.2d at 897. We therefore held that, even assuming the validity of the Baldus study's methodology, "the statistics are inadequate to entitle McCleskey to relief on his constitutional claim." Id. at 896. Our holding in McCleskey forecloses any possible reliance on the Baldus study in the instant case, and appellant is not entitled to an evidentiary hearing on the Baldus study.

With respect to the Bowers and Pierce study, we conclude that appellant's proffer fails to demonstrate a reasonable possibility that the study might compel an inference of purposeful discrimination. The Bowers and Pierce study purported to show a disparity in the imposition of the death penalty in Georgia and other states, based on race, between July, 1972 and August, 1980. Unlike the Baldus study, however, the Bowers and Pierce study did not attempt "to control for all of the factors which play into a capital crime system." Id. at 887. In fact, the only factor controlled for in the Bowers and Pierce study was whether the homicide was a felony murder or non-felony murder. See Bowers & Pierce, Arbitrariness and Discrimination under Post-Furman Capital Statutes, 26 Crime & Delinq. 563, 597-600 & Table 3 (1980). On this basis alone, we find the Bowers and Pierce study legally insufficient to compel an inference of purposeful discrimination in the imposition of the death penalty.

In addition, we note that the same Bowers and Pierce study was proffered in Smith v. Balkcom, 671 F.2d 858, 859 (5th Cir. Unit B 1982). There, a panel of this court identified problems with the study that cast serious doubt on its overall validity. The Smith v. Balkcom panel stated:

Appellant's statistician sought to determine the total number of incidents involving homicide reported as having taken place in Georgia by a somewhat arbitrary (but accepted as statistically correct) adjustment for unreported incidents. He used Supplemental Homicide Reports (SHRs) submitted by Georgia law enforcement officers to the Federal Bureau of Investigation. (He compiled a supplemental data base of those homicides associated with the commission of some other felony--one, but only one, of the aggravating circumstances under the statute....) The study then compares these reported incidents with death penalties ultimately imposed, after trial, in the state. No data is offered as to whether or not charges or indictments grew out of reported incidents or as to whether charges were for murder under aggravating circumstances, murder in which no aggravating circumstances were alleged, voluntary manslaughter, involuntary manslaughter, or other offenses. The data are not refined to select incidents in which mitigating circumstances were advanced or found or those cases in which evidence of aggravating circumstances was sufficient to warrant submission of the death penalty vel non to a jury. No incidents resulting in not guilty verdicts were removed from the data. The unsupported assumption is that all such variables were equally distributed, racially, sexually, offender and victim, throughout the SHRs. No conclusions of evidentiary value can be predicated upon such unsupported assumptions.

671 F.2d at 860 n. 33 (citations omitted).

In short, the Bowers and Pierce study shares none of the characteristics that made the Baldus study worthy of extensive judicial scrutiny. Because appellant failed to demonstrate a reasonable possibility that the Bowers and Pierce study might compel an inference of purposeful discrimination in the imposition of the death penalty, we hold that appellant is not entitled to an evidentiary hearing on the Bowers and Pierce study.12

Finally, with respect to the Dix study, appellant proffered the study solely for the purpose of demonstrating that the Georgia Supreme Court's appellate review of death penalty cases does not eliminate the alleged arbitrariness and racial discrimination in the Georgia capital sentencing system. The appellant did not claim, nor could he have claimed, that the Dix study showed that appellate review produced discrimination or increased the level of discrimination in the system.13 In the absence of independent, credible evidence of purposeful discrimination, we find it irrelevant whether or not the Georgia Supreme Court's appellate review procedure is adequate to eliminate such discrimination. We hold that the Dix study, standing alone, does not provide a basis for habeas corpus relief, and appellant is not entitled to an evidentiary hearing on the Dix study.

In conclusion, appellant has failed to proffer any evidence that might compel an inference of purposeful discrimination in the imposition of the death penalty in Georgia. We therefore affirm the district court's denial of appellant's request for an evidentiary hearing as to the arbitrary and racially discriminatory application claim.

Based on the foregoing discussion, we hereby AFFIRM in part the district court's denial of Ross' petition, but we REMAND to the panel for further consideration Ross' claim relating to the jury composition, in light of Ross' motion to supplement the record.

I fully concur in Judge Kravitch's opinion for the court with the exception of Part II.C., which addresses petitioner's claim that the Georgia death penalty statute is applied in an arbitrary and racially discriminatory manner and with inadequate appellate review. As to this claim, I concur in the court's judgment for the reasons I expressed in my concurring opinion in McCleskey v. Kemp, 753 F.2d 877, 904, (11th Cir.1985) (en banc) (Tjoflat, J., concurring).

JOHNSON, Circuit Judge, dissenting in part and concurring in part:

I concur in parts I, II A and II B of the opinion of the Court. I also agree that several of the statistical studies offered by Ross to prove that Georgia has applied its death penalty in an unconstitutionally discriminatory manner were not sufficient to warrant an evidentiary hearing in the district court. I must dissent, however, from the majority's holding that the study conducted by Dr. David Baldus did not demonstrate a "reasonable possibility that the evidence might compel an inference of purposeful discrimination" and therefore did not warrant an evidentiary hearing. The petitioner did not have to prove purposeful discrimination in order to support his claim, and even if he were required to do so, he still would have almost certainly succeeded. In my view, the evidence proffered by Ross would establish grounds for relief under the Eighth Amendment because it would prove that Georgia has applied its death penalty statute in an arbitrary and capricious manner by allowing race to determine in part who will receive the death penalty.

A petitioner can prove that a state has operated its death penalty system arbitrarily and capriciously without proving that state agents intended to discriminate against the petitioner or against any other defendant in a capital case. A death penalty system is arbitrary and capricious if it produces significantly inconsistent results. Godfrey v. Georgia, 446 U.S. 420, 433, 100 S. Ct. 1759, 1767, 64 L. Ed. 2d 398 (1980) (death penalty scheme must provide meaningful way of distinguishing between those who receive death sentence and those who do not). As a result, many of the constitutional limitations on the death penalty focus not on the intent of juries, prosecutors and other state agents, but on the consistency of the results they produce. See Pulley v. Harris, 465 U.S. 37, 104 S. Ct. 871, 79 L. Ed. 2d 29 (1984) (comparative proportionality review not required if system adequately ensures consistent and rational results); Westbrook v. Zant, 704 F.2d 1487, 1504 (11th Cir.1983).

The Court's holding today on the Enmund issue is a classic example of such objective controls on the pattern of outcomes produced by a death penalty system. Regardless of a jury's intent regarding the proportionality of its recommended sentence, the sentence will be upheld if an appellate court can determine that the defendant's crime is consistent with other capital crimes. While some procedural due process requirements designed to ensure the accuracy of the proceedings focus on the actual deliberations of different state actors, see Eddings v. Oklahoma, 455 U.S. 104, 102 S. Ct. 869, 71 L. Ed. 2d 1 (1982); Spivey v. Zant, 661 F.2d 464 (5th Cir. Unit B 1981), cert. denied, 458 U.S. 1111, 102 S. Ct. 3495, 73 L. Ed. 2d 1374 (1982), the violation of a petitioner's substantive rights in the death penalty context may be proven without reference to intent.

An Eighth Amendment claim based on racial discrimination cannot require proof of discriminatory intent because the death penalty heightens the need for consistent and fair decisions while at the same time maintaining the need for the discretion to make individualized judgments. An appellate court can only give searching review to the fairness and consistency of a discretionary decision if it employs evidence of disparate impact because the discretion of state decisionmakers makes evidence of their intent difficult if not impossible to obtain. McCleskey v. Kemp, 753 F.2d 877, 909 (11th Cir.1985) (Johnson, J., dissenting). Since the special nature of the death penalty calls for the use of effects evidence, the evidence proffered by Ross had only to show by a preponderance of the evidence a significant racial influence on the pattern of death sentences; the evidence did not have to compel an inference of purposeful discrimination. The Baldus Study had a reasonable probability of success1 under this standard and the district court erred in refusing to hold a hearing.

Even if proof of purposeful discrimination were the appropriate standard, the Baldus Study proffered by Ross would suffice to support his constitutional claim. In some instances, circumstantial or statistical evidence of racially disproportionate impact may be so strong that it compels an inference of purposeful racial discrimination. Castaneda v. Partida, 430 U.S. 482, 97 S. Ct. 1272, 51 L. Ed. 2d 498 (1977); Smith v. Balkcom, 671 F.2d 858, 859 (5th Cir. Unit B), cert. denied, 459 U.S. 882, 103 S. Ct. 181, 74 L. Ed. 2d 148 (1982). The Baldus Study begins with evidence of a very large disproportionate impact by showing that the death penalty falls on killers of white victims eleven times more often than on killers of black victims. Then, moving beyond the usual confines of statistical evidence, the study analyzes that disproportionate impact and shows it to be strong, virtually irrefutable evidence of purposeful discrimination. The multivariate regression analysis demonstrates that the average killer of a white victim has a significantly greater risk of receiving the death penalty than other killers and that the greater risk exists solely because of the race of the victim. Ross has therefore produced sufficiently strong evidence of a disproportionate impact sufficiently large to compel an inference of purposeful discrimination.

While this Court has already decided, wrongly in my view, the legal insufficiency of the Baldus Study, the issue remains alive because it is still possible that the Supreme Court will take action in McCleskey v. Kemp, supra. The shortcomings of this Court's decision in McCleskey, together with the possibility that it will not stand as binding precedent on this issue, lead me to conclude that the evidence proffered by Ross demonstrates a reasonable possibility of supporting a successful claim for habeas corpus relief. I would reverse the judgment of the district court and order it to hold an evidentiary hearing.

CLARK, Circuit Judge, concurring in part and dissenting in part:

The majority misunderstands Enmund and promulgates a rule that deprives defendants in this circuit of their Eighth and Fourteenth Amendment right to a jury resolution of the fact issue of whether a defendant "killed or attempted to kill, and regardless of whether he intended or contemplated that life be taken...." as well as his Eighth Amendment right not to receive the death penalty unless he killed, attempted to kill or intended that a killing take place. Enmund 102 S. Ct. at 3379. At stake is whether a defendant's right to a jury resolution of the issue of guilt includes the right to a jury determination of the extent of that guilt, if guilt is found.

The majority holds that such a right is subject to proportionality review, indicating a misunderstanding of that principle of law. Proportionality review has been spoken of by courts in two separate ways. First, it has been used with reference to an abstract evaluation of the appropriateness of a sentence for a particular crime. The Supreme Court for example has struck down punishments as inherently disproportionate when imposed for a particular crime or category of crime. See, e.g., Coker v. Georgia, 433 U.S. 584, 97 S. Ct. 2861, 53 L. Ed. 2d 982 (1977) (death penalty disproportionate for the rape of an adult woman); Solem v. Helm, 463 U.S. 277, 103 S. Ct. 3001, 77 L. Ed. 2d 637 (1983) (life sentence without parole disproportionate for a habitual offender whose previous criminal record consisted of seven relatively minor property offenses); and Enmund v. Florida, supra. Second, proportionality review has been used in a comparative sense, i.e., is a penalty that is not per se disproportionate, nevertheless unacceptable in a particular case because it is disproportionate to the punishment imposed on others for the same crime? This type of review is triggered when an individual defendant contends that other defendants convicted of the same crime have been punished less severely than he. See Pulley v. Harris, 465 U.S. 37, 104 S. Ct. 871, 79 L. Ed. 2d 29 (1984).1 This case, and the inquiry suggested by the majority implicates the first type of proportionality review, i.e. inherent disproportionality; the severity of the penalty is constitutionally out of proportion to the nature of the crime.

The Supreme Court cases dealing with inherent disproportionality lead me to a different conclusion from that reached by the majority. More specifically, in Enmund the Supreme Court, pursuant to the Eighth Amendment held that one who "aids and abets a felony in a course in which a murder is committed by others but who does not himself kill, attempt to kill, or intend that a killing take place or that lethal force will be employed" cannot receive the death penalty. 458 U.S. 782, 102 S. Ct. 3368, 3376-77, 73 L. Ed. 2d 1140 (1982). It was the Court's conclusion that the death penalty in such a case was not commensurate with the lesser culpability inherent when the evidence did not establish that the accused intended that a life be taken. Admittedly, Enmund was a proportionality review in the sense that the Court examined the history of felony murder rule statutes in the various jurisdictions and the sentencing patterns of juries in similar situations. However, the conclusion that the Supreme Court reached is not one that fluctuates from case to case but rather was a final determination that one who does not kill, attempt to kill or intend to kill cannot receive the death penalty. To allow the imposition of the death penalty in such a case, the Supreme Court concluded, would be a violation of the Eighth Amendment. Id. 102 S.Ct. at 3372.

Therefore, the determination has been made by the highest tribunal in this country that an "Enmund type" defendant cannot under our Constitution be eligible for the death penalty. The determination of whether an accused is an "Enmund type" defendant is a question of fact. The determination is identical to that of a jury decision of whether an accused is guilty of murder, voluntary manslaughter or involuntary manslaughter.

To receive the death penalty in Georgia, the state must prove that the defendant with malice aforethought caused the death of the victim (O.C.G.A. Sec. 16-5-1(a) and (b) ) or that the defendant in the commission of a felony caused the death of the victim (O.C.G.A. Sec. 16-5-1(c)) (emphasis added). Consideration of this Georgia statute along with the holding in Enmund would require that the evidence show that the defendant "caused" or "intended that a killing take place" and that the jury be so instructed before a defendant could be eligible for the death penalty. The Enmund holding does not limit Georgia's felony murder statute. The conceptual difficulty arises in a case where there is a conspiracy to commit a felony during the course of which a murder takes place. The law of conspiracy makes each member guilty of the criminal conduct of each other member under an agency theory. The holding in Enmund forbids extension of the death penalty to a co-conspirator who has no intention that a life be taken. The same reasoning would apply if the defendant were an aider or abettor. Nothing in Enmund limits the sentence of death in a typical felony murder case or in a conspiracy to commit murder case. It only forbids a death penalty being based on the imputation of an intent to kill. The Georgia cases are clear that a conspirator to the commission of a felony that results in a murder can be convicted of murder,2 but I find no Georgia case where such a defendant, who did not cause a death or intend that a life be taken, received the death penalty.3

In Georgia, the jury is required to make all factual determinations.4 Therefore, the conclusion is inescapable that an Enmund instruction must be given in a case where the evidence warrants it.

This conclusion is not only consistent with but is compelled by the Supreme Court's decisions in capital cases. The central mandate of Furman v. Georgia, 408 U.S. 238, 92 S. Ct. 2726, 33 L. Ed. 2d 346 (1972), has been described by the High Court as being, " '[w]here discretion is afforded a sentencing body on a matter so grave as the determination of whether a human life should be taken or spared, that discretion must be suitably directed and limited so as to minimize the risk of wholly arbitrary and capricious action.' " Zant v. Stephens, 462 U.S. 862, 103 S. Ct. 2733, 2741, 77 L. Ed. 2d 235 (1983). In this vein, the Court has been chiefly concerned with the procedures by which the state imposes the death sentence. Ramos v. California, 463 U.S. 992, 103 S. Ct. 3446, 3451, 77 L. Ed. 2d 1171 (1983). This is so because the qualitative difference of the death penalty requires "a corresponding difference in the need for reliability in the determination that death is the appropriate punishment in a specific case." Woodson v. North Carolina, 428 U.S. 280, 305, 96 S. Ct. 2978, 2991, 49 L. Ed. 2d 944 (1976). Just last term the Court stated, "We reaffirm our commitment to the demands of reliability in decisions involving death...." Spaziano v. Florida, --- U.S. ----, ----, 104 S. Ct. 3154, 3160, 82 L. Ed. 2d 340, 349 (1984).

In Beck v. Alabama, 447 U.S. 625, 100 S. Ct. 2382, 65 L. Ed. 2d 392 (1980), the Court found that the Constitution required that a defendant in a capital case receive a lesser included offense instruction if the facts supported it. The Court concluded "... if the unavailability of a lesser included offense instruction enhances the risk of an unwarranted conviction, Alabama is constitutionally prohibited from withdrawing that option from the jury in a capital case." 447 U.S. at 638, 100 S.Ct. at 2390. This result was reached because of the mandate of the Eighth and Fourteenth Amendments which requires the invalidation of procedural rules that diminish the reliability of the life/death determination. Id. Similarly, a state is prohibited from withdrawing the decision whether an accused is an "Enmund type" defendant from the jury. To allow the states to do so is to enhance the risk of an unwarranted death sentence.

The clear holding of Enmund is that one who does not kill, attempt to kill or intend that a killing take place is not "death eligible." As it is the purpose of a capital statutory scheme to determine who shall receive the death penalty from among those who are "death eligible," any valid trial process must first properly attempt to narrow the pool. This is, for example, the purpose of aggravating circumstances, "they circumscribe the class of persons eligible for the death penalty." Zant, 103 S. Ct. at 2743. Likewise, an Enmund instruction is required to remove from the "death eligible" class those who do not kill, attempt to kill or intend that a killing take place.5 The Eighth Amendment cannot and does not require anything less.

The Court's reasoning in Lockett v. Ohio, 438 U.S. 586, 592, 98 S. Ct. 2954, 2958, 57 L. Ed. 2d 973 (1978), is also useful in this case. In Lockett, the Court held that a state could not preclude the sentencer's "considering as an independent mitigating factor any aspect of a defendant's character or record or any of the circumstances of the offense that the defendant proffers as a basis for a sentence less than death." 438 U.S. at 604, 98 S.Ct. at 2965. The capital sentencing system at issue in Lockett, the Court continued, "... creates the risk that the death penalty will be imposed in spite of factors which may call for a less severe penalty. When the choice is between life and death, that risk is unacceptable and incompatible with the commands of the Eighth and Fourteenth Amendments." Id. at 605, 98 S.Ct. at 2965.

This is a more compelling issue than Lockett. In Lockett, the jury was denied access to information which may in the jury's discretion have resulted in a sentence less than death. The Enmund decision, however, if resolved in the defendant's favor is determinative of eligibility for the life/death decision and therefore the failure to so instruct the jury creates the risk that the death penalty will be imposed despite a factor which constitutionally removes a defendant from the class of those who are death eligible. This is simply not a decision to be left for a reviewing court in the first instance. The risk of error is too great in a capital case.6

The majority's refusal to acknowledge this is to abandon our responsibility to "carefully scrutinize capital sentencing schemes to minimize the risk that the penalty will be imposed in error or in an arbitrary and capricious manner." Spaziano, supra, --- U.S. at ---- n. 7, 104 S. Ct. at 3162 n. 7, 82 L. Ed. 2d at 349 n. 7. Instead of a procedure that removes "Enmund type" defendants from the pool of death eligible persons before the penalty is imposed, the majority places the federal courts in the position of making a post-trial determination that a particular accused was or was not within the parameters of the Court's holding in Enmund.

The substantive limitation placed upon capital sentencing by the Eighth Amendment in Enmund prohibits the imposition of the death penalty upon one who does not kill, attempt to kill or intend that a killing take place. The procedural component of the Eighth Amendment requires that the discretion of the jury be focused upon the individual's culpability by an appropriate instruction. Contrary to the majority, I would conclude that not only is an Enmund instruction useful or necessary, but that it is Constitutionally required. This conclusion is consistent with that reached by the Fifth Circuit in Reddix v. Thigpen, 728 F.2d 705 (5th Cir.1984), and Clark v. Louisiana State Penitentiary, 694 F.2d 75 (5th Cir.1982).7

The majority reaches out to formulate a holding to deny "Enmund type" defendants a jury determination of an issue that determines whether he is elgible to even be considered for a possible penalty of death. It arrogates to the federal courts in habeas corpus cases the authority to conduct a case by case proportionality review in death cases, an authority we do not have.8 In this case, sentence review is not even implicated. An Enmund issue is directed to determining whether an accused charged with first degree or felony murder is guilty to the extent that he should be subjected to death penalty consideration by a jury and/or judge. The three states in this circuit grant a person charged with murder a bifurcated trial to determine first whether the accused is guilty (the guilt phase) and should be considered for the death penalty, and if so, a secondary trial on the issue of whether the sentence is death or a life sentence (the sentencing phase).9

While I completely disagree with the rationale of the majority opinion, I do agree that the judgment of the district court should be affirmed. Ross was not entitled to an Enmund instruction.10 The majority opinion reviews the evidence and I shall not repeat it. There is no doubt that Ross contemplated that a life would be taken if a policeman entered the house, and there is only a whimsical doubt as to whether Ross was the actual killer. He either killed Meredith or his gun misfired while he was trying to kill him. Even accepting the latter version, Enmund is merely a straw-man in this case.

Therefore, I would affirm the district court's denial of the writ insofar as this particular issue is concerned. I concur with the remaining portions of the opinion.