The opinion of the court was delivered by: Padova, J.
(DEATH PENALTY HABEAS CORPUS)
Before the Court for the second time is Robert Lark's Petition for Writ of Habeas Corpus filed pursuant to 28 U.S.C. § 2254. On July 3, 2007, we conditionally granted the Petition, finding that the Commonwealth of Pennsylvania failed to satisfy its burden at the second step of the analysis established by the United States Supreme Court in Batson v. Kentucky, 476 U.S. 79 (1986).*fn1 See Lark v. Beard, 495 F. Supp. 2d 488, 503 (E.D.Pa. 2007) ("Lark I") vacated sub nom Lark v. Secretary, Pa. Dept. of Corr., 645 F.3d 596 (3rd Cir. 2011) ("Lark II"). On June 16, 2011, the United States Court of Appeals for the Third Circuit vacated our Order granting the Petition. Lark II, 645 F.3d at 629. The Commonwealth did not challenge our finding that Lark established a prima facie case at step one, but confined its challenge to our holding that Lark was entitled to the entry of the conditional writ of habeas corpus because the Commonwealth had not articulated a race-neutral explanation for its strikes at step two. Id. at 620-21. The Third Circuit determined that, notwithstanding the Commonwealth's failure to meet its step two burden of demonstrating a race-neutral ground for its peremptory strike of a minority veniremen, the burden remains on Petitioner to show by a preponderance of the evidence at step three that the prosecutor engaged in purposeful discrimination. Lark II, 645 F.3d at 625-27 (analyzing Johnson v. California, 545 U.S. 162 (2005) and explaining that the prosecutor's lack of response at step two is evidence to be taken into account at step three, but is not, by itself, of such dispositive force that it establishes that there was a Batson violation). Presently before the Court are the parties' additional submissions on the step three issues. We find that Petitioner has satisfied his burden of demonstrating by a preponderance of the evidence that the prosecutor engaged in purposeful discrimination in striking African-American veniremen from his jury. Accordingly, we again conditionally grant the writ.
I. FACTUAL BACKGROUND AND PROCEDURAL HISTORY
The Third Circuit fully set forth the factual basis for Lark's conviction and the direct and collateral appeal history in Lark II, 645 F.3d at 599-606, which we incorporate by reference herein. To summarize, Petitioner was convicted of first degree murder, possession of an instrument of crime, terroristic threats, and kidnaping in the Court of Common Pleas of Philadelphia County on June 28, 1985, and was subsequently sentenced to death. The convictions resulted from Lark's killing of Tae Bong Cho, and his kidnaping and restraint of Cho's two young children and their mother. Petitioner appealed, and the Supreme Court affirmed his convictions and sentences on May 20, 1988. See Commonwealth v. Lark, 543 A.2d 491 (Pa. 1988).
Following extensive state collateral proceedings, Lark filed the instant petition for writ of habeas corpus, pursuant to 28 U.S.C. § 2244, on March 16, 2001, alleging an Eighth Amendment ineffective assistance of counsel claim in connection with both the guilt and penalty phases of his trial, as well as a Fourteenth Amendment Equal Protection claim. Lark II, 645 F.3d at 602 n.9. After determining that Petitioner alleged facts that, if true, would satisfy his Batson step one burden by making a prima facie showing that the prosecutor exercised a peremptory challenge on the basis of race, we conducted an evidentiary hearing. Id. at 603.
To support its step two burden, we permitted the Commonwealth to present circumstantial evidence to reconstruct the prosecutor's peremptory strike decision-making process. The prosecutor, having first testified that he had no current recollection of the Lark jury selection, nor any independent or refreshed recollection of the voir dire after having read the voir dire transcript, Lark I at 495, was allowed to testify to his practice of considering certain criteria in exercising peremptory challenges. He identified the following factors that he routinely considered and the reason he considered that factor significant: (1) the neighborhood where the potential juror lived was significant to his decision to exercise a peremptory strike because he did not want jurors who lived near the defendant or near to where the incident occurred (N.T. 11/8/06 at 73-74); (2) the potential juror's employment status, because persons with a job have roots in the community (id. at 82); (3) the potential juror's age, because older jurors are wiser and more responsible (id. at 94); (4) whether the potential juror had relatives that were police officers, which he considered a positive factor (id. at 95); (5) the potential juror's home ownership, because it showed a stake in the community (N.T. 11/9/06 at 10); (6) vocation -- he did not want teachers and social workers as jurors (id.); (7) hardship (id. at 48); (8) prior jury experience (id.); (9) whether the juror had been a victim of crime or a witness or defendant in a criminal case (id.); (10) whether the juror knew any potential witness in the trial (id.); (11) whether the juror had any feelings about the death penalty (id.); and (12) whether the potential juror had children the same age as the defendant (N.T. 11/8/06 at 102). The prosecutor denied striking any juror because of his or her race. (N.T. 11/9/06 at 51.) He testified that using race as a factor in jury selection was legally and morally abhorrent to him. (N.T. 11/8/06 at 71-72.)
The prosecutor also testified as to why he thought he may have stricken certain members of the venire, based upon his review of the voir dire transcript. However, even after he reviewed the transcript, the prosecutor could state no reason for striking three African-American veniremen: Shirley Sampson, Florence Williams, and Edison Sisco. Lark II, 645 F.3d at 604-05. We determined in Lark I that the Commonwealth's failure to satisfy its burden of production with respect to the prosecutor's peremptory strikes of Sampson, Williams, and Sisco at step two of the Batson analysis mandated the granting of the writ. Lark I at 503. The Third Circuit agreed that the Commonwealth failed to satisfy its burden of production at step two, Lark II, 645 F.3d at 603; however, the Third Circuit disagreed with our conclusion that the Commonwealth's failure at step two warranted granting the writ. Id. at 626-28. The Third Circuit indicated that, where the passage of time has diminished the prosecutor's recollection of the voir dire, the district court should proceed to the third step of the Batson analysis. Id. at 628.
II. STEP THREE OF THE BATSON ANALYSIS
In Williams v. Beard, 637 F.3d 195 (3d Cir. 2011), the Third Circuit restated the law applicable to step three:
At step three of the Batson analysis, the petitioner must show that "it is more likely than not that the prosecutor struck at least one juror because of race." Bond [v. Beard, 539 F.3d 256 (3d Cir. 2008)] at 264. To determine whether the petitioner has carried his or her burden, the court must evaluate "all evidence introduced by each side (including all evidence introduced in the first and second steps) that tends to show that race was or was not the real reason" for each strike. Hardcastle [v. Horn, 368 F.3d 246 (3d Cir.2004)] at 259 (quoting Riley v. Taylor, 277 F.3d 261, 286 (3d Cir. 2001) (en banc)); see also Snyder v. Louisiana, 552 U.S. 472, 478, 128 S.Ct. 1203, 170 L.Ed.2d 175 (2008) (explaining that "all of the circumstances that bear upon the issue of racial animosity must be consulted"). Step three ultimately focuses upon the prosecutor's subjective motivation, which ideally includes an assessment of the demeanor and credibility of the various voir dire participants. See Snyder, 552 U.S. at 477, 128 S.Ct. 1203 ("Step three of the Batson inquiry involves an evaluation of the prosecutor's credibility, and the best evidence [of discriminatory intent] often will be the demeanor of the attorney who exercises the challenges." (alteration in original) (internal citations and quotation marks omitted)).
Williams, 637 F.3d at 215-16. As this discussion makes clear, a key part of our analysis at step three will be determining whether or not we accept the prosecutor's assertion that race played no part in his decision making process. In making this determination, we consider the entire record, including the statistical evidence Petitioner introduced at step one, see id. at 214 (noting that statistical evidence may be sufficient by itself to make out a prima facie case of racial discrimination at step one, citing Jones v. Ryan, 987 F.2d 960, 971 (3d Cir. 1993)), as well as the evidence offered by the Commonwealth at step two, to show what factors the prosecutor ordinarily considered in making peremptory strike decisions. "An explanation that appears race neutral at step two may betray an improper motive if it is invoked to strike African Americans but not other non-black venirepersons exhibiting the same characteristic." Id. at 216.
At step three, we also examine "comparator" evidence, "comparing stricken members of the venire to individuals the Commonwealth deemed acceptable." See id. ("We have previously authorized such an evaluative procedure, explaining, 'A comparison between a stricken black juror and a sitting white juror is relevant to determining whether the prosecution's asserted justification for striking the black juror is pretextual.'") (quoting Riley, 277 F.3d at 282; Holloway v. Horn, 355 F.3d 707, 724 (2004)); see also Snyder, 552 U.S. at 479--86, 128 S.Ct. 1203 (performing comparative analysis); Miller--El v. Dretke, 545 U.S. 231, 241--52 (2005) ("Miller-El II") (finding Batson violation based in part on juror comparison evidence).
The Third Circuit in Williams stated that, when conducting a comparison analysis, it is insufficient to proffer venire members who lack one or more of the characteristics upon which the prosecutor exercised a strike. This is not to erect an unreasonable roadblock; rather, it ensures accuracy in an area often guided by guesswork and hunches. See [United States v.] DeJesus, [347 F.3d 500 (2003)] at 505 (explaining that a peremptory challenge "is usually based on educated guesses about probabilities based on the limited information available to an attorney about prospective jurors"). Because the focus in step three is to uncover a prosecutor's subjective motivation, it is imperative to account for the complete combination of factors that caused the prosecutor to exercise a strike.
Williams, 637 F.3d at 217. The comparator evidence in Williams was held to be unpersuasive, however, because the petitioner could identify no comparators that met all of the criteria exhibited by the stricken veniremen. Id. However, as the Supreme Court explained in Miller-El II, there is no per se rule "that no comparison is probative unless the situation of the individuals compared is identical in all respects, and there is no reason to accept one. . . . A per se rule that a defendant cannot win a Batson claim unless there is an exactly identical white juror would leave Batson inoperable; potential jurors are not products of a set of cookie cutters." Id. at at 247 n.6; see also Snyder, 552 U.S. at 483 (finding that the prosecutor's acceptance of white jurors who disclosed information "at least as serious" as that disclosed by stricken African-American juror made prosecutor's reliance on that factor implausible).
Petitioner argues, based upon the trial record of the voir dire and the evidence adduced at our evidentiary hearing, that it is more likely than not that the Commonwealth engaged in purposeful discrimination in the exercise of peremptory strikes against African-American veniremen. To support his argument Petitioner relies upon (a) a statistical analysis of the prosecutor's peremptory strikes, (b) comparisons among jurors, and (c) the prosecutor's comments when Petitioner's attorney objected to his pattern of peremptory strikes.*fn3
1. Strike rates and acceptance rates
The parties stipulated to the racial composition of the venire from
which Petitioner's jury was selected. Lark II, 645 F.3d at 605-06.
After all veniremen excused for cause or by consent*fn4
excluded, the jury, which totaled 14 people, including the two
alternates, was selected from a venire consisting of 45 people. Of
those 45 people, the prosecutor had the opportunity to accept or
strike 29 veniremen before the jury was filled. Of these 29 people, 17
were African-American and 12 were Caucasian. (App. 1.) The prosecutor
exercised 15 peremptory strikes. (Id.) Of those 15 strikes, 13 were
used to strike African-American veniremen -- resulting in an 87%
strike rate. (Id.; Lark II at 606). He used two strikes against
Caucasian veniremen -- a 13% strike rate. (Id.)*fn5
The prosecutor accepted only four of the 17 African-American jurors he
had the opportunity to accept -- for an acceptance rate of 23.5%; he
accepted 10 of the 12 Caucasian jurors he had the opportunity to
accept -- for an acceptance rate of 83%. (Id.)
The Commonwealth argues that this evidence does not establish at step three that any particular strike was the product of discrimination. First, it notes that four African-Americans were seated as jurors, which, it argues, lends credibility to the prosecutor's testimony that he was not excluding jurors because they were African-American. (Resp. at 3-4 citing DeJesus, 347 F.3d at 509 (stating that the fact that one Hispanic and three African Americans were seated in the final jury, while the government had three unused three peremptory strikes, makes the government's race-neutral explanation more believable)).
The Commonwealth also asks us to consider another statistic -- the prosecutor's acceptance rate -- which it contends is more "benign." (Resp. at 4.) As noted, of the 17 African-Americans that the prosecutor had the opportunity to accept or strike, he accepted four, or about 23.5%. We consider the African-American acceptance rate, along with all of the other evidence, as relevant to the step three determination. However, a comparison with the prosecutor's acceptance rate for Caucasian jurors, 10 out of 12 or 83%, is also evidence we consider.
Third, the Commonwealth notes that there is no evidence concerning the race of the entire venire or the race of jurors struck by the defense. It contends that, in the absence of this evidence there is no way to accurately calculate the overall exclusion rate, i.e., to compare the racial makeup of the entire venire with the racial makeup of the seated jury. See United States v. Gooch, 665 F.3d 1318, 1327, 28 (D.C. Cir. 2012) (stating "When the 'jury composition mirror[s] the make-up of the venire' and 'the prosecutor's strikes [do] not skew the racial composition of the resulting jury,' '[t]he circumstances . . . are a far cry from the facts of cases in which the Supreme Court has found a Batson violation.'" (quoting United States v. Moore, 651 F.3d 30, 41 (D.C. Cir. 2011) (per curiam))). It asserts that Petitioner's failure to establish the racial makeup of the entire venire is an independent reason why his Batson claim fails because the exclusion rate is necessary for reliable assessment of the statistical evidence. (Resp. Mem. at 5 (citing Lewis v. Horn, 581 F.3d 92, 103 (3d Cir. 2009)).
The Commonwealth's reliance on Lewis is, however, misguided. The Lewis court was discussing Batson's step one requirements, not step three, and Lewis's claim failed because he had no evidence to support his allegations that 66.6% of the members of the venire stricken by the prosecutor were African-American. Id. at 103-04. The Lewis Court also stated that the exclusion rate is not absolutely necessary to satisfy Batson's step one requirements. Lewis at 103 (stating "In Abu-Jamal [v. Horn, 520 F.3d 272 (3d Cir. 2008)] we emphasized the importance (although not necessity) of supplying information about the strike rate and the exclusion rate in order to demonstrate a prima facie violation of Batson") (parenthetical in original). Moreover, we are required when we determine whether the Petitioner has satisfied his burden at step three to evaluate "all evidence introduced by each side (including all evidence introduced in the first and second steps) that tends to show that race was or was not the real reason" for each strike. Williams at 215 (quoting Hardcastle, 368 F.3d at 259). Focusing solely on the failure to establish the "overall" exclusion rate would violate this principle.
The Commonwealth also argues that Petitioner's strike and acceptance statistics evidence fails to meet his step three burden because in four recent cases, all decided after we issued Lark I, the Third Circuit rejected Batson claims that cited similar statistical evidence. First, we reject the inference that these cases mandate that Petitioner's statistical evidence should be viewed in isolation from all other material evidence at step three. As just stated, we must look to the totality of the evidence to determine whether race was or was not the real reason for each strike. Second, we find that none of the four cases mandate a conclusion that it is more likely than not that Petitioner cannot demonstrate racial discrimination in jury selection in this case.
The first case on which the Commonwealth relies is Williams, 637 F.3d 195. In Williams, the prosecutor accepted five of the 19 African Americans she had the opportunity to strike, for an acceptance rate of 26.3%. By contrast, her acceptance rate of white veniremen was 90%. Id. at 215. The Third Circuit found that the variation in the prosecutor's acceptance rates for African-American and Caucasian veniremen was sufficient to establish a prima facie showing of racial discrimination at step one. Id. However, the Court concluded that the comparison evidence presented by the petitioner was insufficient to establish that the prosecutor's race-neutral explanations for her strikes were pretextual. Id. at 216-19. The Third Circuit concluded that Williams's Batson challenge failed on its merits at step three because the totality of petitioner's proofs -- not just the statistics -- did not overcome the prosecutor's stated non-discriminatory reasons for her strikes. Id. at 219-220.
Similarly, in the second decision cited by the Commonwealth, Lewis, 581 F.3d 92, the petitioner claimed that the prosecutor exercised eight of his 12 peremptory strikes against African-Americans, for a strike rate of 66.67%. Id. at 104. Further, petitioner alleged he was tried by an all-Caucasian jury. Id. The Third Circuit found that petitioner had not made out a step one prima facie case of discrimination because he had not cited to any record evidence or any other support outside the record, to substantiate his bare allegations. Id. (stating "Without information about the number and racial composition of the entire venire, we cannot calculate the exclusion rate and we lack the 'contextual markers' to analyze the significance of the strike rate."). The Court further stated that, even if it accepted as true "Lewis's bald assertion that eight of the twelve venire members whom the prosecutor struck were African-American, a strike rate of 66.67% is insufficient information to establish a prima facie case of racial discrimination in the exercise of peremptory strikes." Id. Here, in contrast, Petitioner identified the race of the 29 veniremen for whom the prosecutor exercised a choice, the prosecutor's strike rate of African-Americans was significantly greater than 66.67%, and Petitioner has presented other evidence on the step three issue. Thus, both Williams and Lewis have little application to this case.
Next, the Commonwealth relies upon Abu-Jamal, 520 F.3d 272. In Abu Jamal, the prosecution strike rate of African-Americans was 66.67% with three African-Americans chosen as jurors. Id. at 287, 291. While the Commonwealth asserts that the decision "emphasized the difficulty of evaluating the significance of the strike rate without knowing the racial makeup of the venire, from which the exclusion rate might be calculated," (Resp. Mem. at 7 (emphasis in original) (citing Abu-Jamal, 520 F.3d at 290-92)), this was not the import of the opinion. In a decision governed by the AEDPA, the Third Circuit determined the state court rejection of the Batson claim was not unreasonable in part because petitioner had the opportunity to question the prosecutor about his striking African-American veniremen at a hearing in state court and failed to do so. Id. at 292. Because Abu Jamal failed to adequately develop the record when he had that opportunity, the Third Circuit concluded that the state court finding that he had not made out a prima facie case at step one was a reasonable interpretation of Batson. Id.
Lastly, the Commonwealth relies upon Bond, 539 F.3d 256. In Bond, the prosecution strike rate of African-Americans was 78.6% with four African-Americans chosen as jurors (two were alternates). Id. at 269. The Commonwealth asserts that this case stands for the proposition that "raw statistics do not provide a clear picture of intentional racial discrimination." (Resp. Mem. at 7 (quoting Bond, 539 F.3d at 269)). While this quote is correct, it has been taken out of context. In fact, while the Third Circuit was acknowledging that this statistic was mitigated by the fact that the percentage of African-Americans on the jury was actually higher than the percentage of African-Americans in one of the two venire panels considered by counsel during Bond's jury selection, that was not the sole basis for the Bond Court's conclusion that the petitioner had failed to meet his obligations under step three. Id. at 269-275. In Bond, the prosecutor offered contemporaneous non-discriminatory reasons for his strikes and the trial court made findings that these were not pretextual. Id. at 271-72. On habeas review, the Third Circuit afforded deference to these holdings.
Here, we deal in large part with stricken jurors for whom the prosecutor could identify no non-discriminatory reasons for his strikes apparent from the voir dire transcript. Because none of the four decisions the Commonwealth cites mandates a conclusion that Petitioner's statistical evidence should not be reviewed in connection with his step three burden, we treat the prosecutor's strike rate and acceptance rate statistics as part of the totality of evidence tending to show whether or not race was the real reason for each strike.
2. Analysis of the prosecutor's jury selection criteria To support its step two burden, we permitted the Commonwealth to present circumstantial evidence to reconstruct the prosecutor's peremptory strike decision-making process. This evidence is also relevant at step three. Hardcastle, 368 F.3d at 259. In addition to relying upon the fact that the prosecutor could state no reason whatsoever for striking at least three African-American veniremen, Petitioner argues that our factual findings, along with the testimony of record from the voir dire, support the step three conclusion that the prosecutor engaged in purposeful discrimination because the record demonstrates that the prosecutor applied the factors ...