Rodney E. Thomas (Thomas) was convicted by a Kern County, California jury of rape, sodomy, oral copulation, and false imprisonment. He now appeals the denial of his 28 U.S.C. § 2254 petition for writ of habeas corpus, alleging that he was deprived of his Sixth and Fourteenth Amendment rights to a jury drawn from a representative cross section of the community, and that his trial counsel provided him with ineffective assistance by failing to raise, investigate, or preserve a challenge to the composition of his jury panel.

FACTS AND PROCEEDINGS BELOW

Thomas, a black male, was arrested in December 1983 and charged with one count of false imprisonment, three counts of oral copulation, two counts of rape, and two counts of sodomy in connection with an alleged sexual assault upon a woman in Kern County, California. On April 26, 1984, Thomas’ trial began in Kern County Superior Court before an all-white jury. Thomas’ trial counsel did not raise any challenge to the composition of the jury panel before the jury was sworn. See Cal.Code Civ. Proc. § 225 (requiring that any challenge to the composition of the jury panel be made prior to the swearing of the trial jury). Thomas’ defense at trial was that he engaged in consensual sex with the victim. On May 16, 1984, the jury convicted Thomas on all counts, and he is currently serving a 59-year sentence in state prison.

Thomas appealed to the California Court of Appeal, which affirmed his conviction on March 25, 1986. With respect to his fair cross-section claim, the Court of Appeal held that Thomas failed to show that the master jury list in Kern County in 1984 underrepresented blacks. Thomas then filed a state petition for writ of habeas corpus, alleging ineffective assistance of trial counsel, which was denied by the California Supreme Court on October 26,1988.

In December 1988, Thomas filed a federal habeas petition, alleging ineffective assistance of both trial and appellate counsel, and deprivation of his right to a jury drawn from a fair cross section of the community. The district court granted Thomas’ counsel permission to conduct discovery, including deposing the Kern County Jury Commissioner and requiring her to produce master jury lists and specific venire and jury lists for April and May of 1984. The district court also approved discovery from census agencies of statistical evidence regarding the racial composition of the Kern County population during the relevant period.

On December 22, 1994, the district court adopted in full a magistrate judge’s findings and recommendations, denying Thomas’ motion for summary judgment and granting the State’s motion for summary judgment on Thomas’ fair cross-section claim. Regarding Thomas’ ineffective assistance of counsel claims, the district court ordered further briefing. On September 25,1995, the district court adopted in full the magistrate judge’s findings and recommendations regarding the ineffective assistance of counsel claims, denying Thomas’ motion for summary judgment, and denying his petition for writ of habeas corpus. Thomas now appeals the denial of his petition and we affirm.

DISCUSSION

Thomas alleges that he was deprived of his Sixth and Fourteenth Amendment rights to a trial by a jury drawn from a representative cross section of the community. He further alleges that his trial counsel provided him with ineffective assistance by failing to raise, investigate, or preserve a challenge to the composition of the jury panel.

Thomas, a black male, contends that his conviction must be reversed because he was convicted by an all-white jury selected from a panel from which black persons had been systematically excluded. At his sentencing hearing before the state trial court, Thomas complained that the panel of jurors from which his trial jury was chosen contained no black men or women and at most seven Hispanics.

“We review independently and non-deferentially a challenge to the composition of grand and petit juries.” United States v. Sanchez-Lopez, 879 F.2d 541, 546 (9th Cir.1989) (citation omitted). We apply federal law because this case arises under the federal habeas statute and Thomas’ underlying claim is that he was deprived of federal constitutional (Sixth and Fourteenth Amendment) rights. See Bittaker v. Enomoto, 587 F.2d 400 , 402 n. 1 (9th Cir.1978) (“Federal courts must apply federal constitutional law in cases properly before them under the federal habeas statute.”) (citing Sesser v. Gunn, 529 F.2d 932, 934-35 (9th Cir.1976)).

The Supreme Court has developed the following three-part test for establishing a prima facie violation of the Sixth Amendment’s fair cross-section requirement:

Duren v. Missouri, 439 U.S. 357, 364 , 99 S.Ct. 664 , 58 L.Ed.2d 579 (1979); If the defendant makes a prima facie showing under Duren , the burden shifts to the state to justify the infringement by demonstrating that attainment of a fair cross section is incompatible with a significant state interest. Id. at 367-68 , 99 S.Ct. 664 .

In order to establish a prima facie Fourteenth Amendment equal protection violation in the jury selection process, a defendant also must show discriminatory intent. See United States v. Esquivel, 88 F.3d 722, 725 (9th Cir.), cert. denied, - U.S. -, 117 S.Ct. 442 , 136 L.Ed.2d 339 (1996) (citing Castaneda v. Partida, 430 U.S. 482, 494 , 97 S.Ct. 1272 , 51 L.Ed.2d 498 (1977)). Thomas has not even attempted to make such a showing here.

Rather, Thomas argues that he has established his prima facie case under Du-ren. Because there is no dispute that blacks are a distinctive group for purposes of Sixth Amendment analysis, Thomas has satisfied the first prong of the Duren test. However, the district court held that Thomas failed to satisfy the second prong, and we agree.

To meet the second prong of Duren , Thomas must show that the representation of blacks in Kern County venires from which juries were selected was not fair and reasonable in relation to the number of blacks in the community. See Sanchez-Lopez, 879 F.2d at 547 . We have held that “[t]he second prong of the Duren test requires proof, typically statistical data, that the jury pool does not adequately represent the distinctive group in relation to the number of such persons in the community.” Esquivel, 88 F.3d at 726 . In determining the underrepre-sentation of a particular group in jury veni-res, we use an absolute disparity analysis. “We determine absolute disparity by taking the percentage of the group at issue in the total population and subtracting from it the percentage of that group that is represented on the master jury wheel.” Sanchez-Lopez, 879 F.2d at 547 ; see also Esquivel, 88 F.3d at 726 .

California courts have used several different tests to determine the degree of under-representation of minorities on jury venires. See Bell, 49 Cal.3d at 527 n. 14, 262 Cal.Rptr. 1 , 778 P.2d 129 (describing the different tests, but declining to adopt a particular one). Thomas specifically cites the comparative disparity test, which requires the court to divide the absolute disparity percentage by the percentage of the distinctive group in the total population. However, the comparative disparity test is strongly disfavored in the Ninth Circuit on the ground that it exaggerates the effect of any deviation. See Sanchez-Lopez, 879 F.2d at 548 (citation omitted). Because Thomas brings his present fair cross-section claim in a federal habeas proceeding, we apply this circuit’s absolute disparity test in resolving his claim. See Bittaker, 587 F.2d at 402 n. 1.

Thomas is unable to provide the statistics necessary for this court to determine the absolute disparity in his case. The only evidence presented by Thomas regarding the representation of blacks on Kern County jury venires or panels at the time of his trial is his observation that there were no blacks on the particular panel from which his trial jury was chosen. However, the prospective jurors in Thomas’ courtroom represented only a small portion of the venires in April of 1984, when Thomas’ jury was selected, and of the master jury list from which the jurors were selected. Thomas presents no evidence regarding the composition of Kern County venires or panels over time, or of the composition of the master jury wheel.

There is no dispute that part of the reason for Thomas’ inability to provide us with the requisite statistics is his trial counsel’s failure to challenge the composition of the jury panel prior to trial, coupled with counsel’s failure to preserve the relevant statistics at the time of trial. Because Kern County keeps records of venires for only three years, the statistics for April of 1984 no longer exist. We address the. constitutional implications of counsel’s failings below, see discussion infra Part II.; however, for purposes of our fair cross-see tion analysis, the reason for Thomas’ lack of evidence is immaterial. Because Thomas has provided us with insufficient statistical evidence to determine whether blacks were substantially underrepresented on jury venires or panels in Kern County at the time of his trial, his Sixth Amendment fan1 cross-section claim must be denied.

Furthermore, even if we use Thomas’ observation regarding the composition of only his particular jury panel, the absolute disparity in his case does not rise to the level of a Sixth Amendment violation. Relying on the California Court of Appeal’s decision in People v. Alexander, 163 Cal.App.3d 1189, 1199 , 210 Cal.Rptr. 306 (1985), Thomas cites federal census figures showing that the general black population of Kern County was approximately five percent (5%) and the general Hispanic population was approximately eighteen percent (18%) in 1980. Thomas also cites evidence that by 1984, the population figures for blacks and Hispanics had risen to 5.135% and 23.413%, respectively. Using these numbers, the calculation for absolute disparity in this case, where there were zero (0) blacks on the jury panel, is approximately five percent (5%) [5%-0%]. We have previously rejected claims of substantial underrepresentation where the absolute disparity was close to or even greater than this. See Esquivel, 88 F.3d at 727 (4.9%); United States v. Suttiswad, 696 F.2d 645, 649 (9th Cir.1982) (7.7%). Thus, the disparity in this case is insufficient to make out a prima facie ease under the Sixth Amendment.

Alternatively, the magistrate judge, in his September 22, 1994 findings and recommendations, calculated absolute disparity based on Thomas’ 1984 census figures and the jury venire figures used in Alexander . Assuming these are the appropriate statistics, the absolute disparity is still insufficient to make out a prima facie case of a Sixth Amendment violation. If, in 1984, 5.135% of the population of Kern County was black, and Alexander can be relied upon for the proposition that 3.2% of the jurors on Kern County venires were black, the absolute disparity is only 1.94%. As noted above, we have previously rejected claims of substantial un-derrepresentation where the absolute disparity was greater than this. See Esquivel, 88 F.3d at 727 (4.9%); Sanchez-Lopez, 879 F.2d at 547 (2.05%); Suttiswad, 696 F.2d at 649 (7.7%); United States v. Armstrong, 621 F.2d 951, 955-56 (9th Cir.1980) (2.83%); United States v. Kleifgen, 557 F.2d 1293, 1296-97 (9th Cir.1977) (2.9%); US. v. Potter, 552 F.2d 901, 905-06 (9th Cir.1977) (2.7%).

II.

Thomas’ next and related Sixth Amendment claim is that his trial counsel was ineffective in failing to raise the issue of under-representation of blacks prior to trial, and in failing to obtain and preserve records and statistics regarding the 1984 master jury list and jury venires in other cases tried around the same time as Thomas’ trial. According to Thomas, at the time his trial jury was being selected, a reasonably effective California criminal defense lawyer would have known that one of the “most hotly debated issues in California” involved underrepresen-tative juries for black defendants. Faced with a black defendant charged with a sex crime, where the defense was consent, counsel should have been sensitive to. any racial issues. Thus, Thomas argues that defense counsel should have challenged the Kern County jury selection process when he saw that there were no blacks on Thomas’ jury panel.

In order to succeed on an ineffective assistance of counsel claim, a defendant first must demonstrate that counsel’s performance was deficient, that is, that it fell below an objective standard of reasonableness. Strickland v. Washington, 466 U.S. 668, 687 , 104 S.Ct. 2052 , 80 L.Ed.2d 674 (1984). Even if a defendant can establish that counsel’s performance was constitutionally deficient, he still must show that he was prejudiced by counsel’s errors. Id. We need not evaluate both prongs of the Strickland test if the defendant fails to establish one. Pollard v. White, 119 F.3d 1430, 1435 (9th Cir.1997).

Here, we do not address the first prong of the test because, even assuming that Thomas could demonstrate that his counsel’s performance was constitutionally deficient, he has failed to show that he suffered prejudice as a result of counsel’s alleged errors. In order to satisfy the prejudice prong, a defendant must establish that there is a reasonable probability that, but for counsel’s errors, the result of the proceeding would have been different. See Strickland, 466 U.S. at 693-94 , 104 S.Ct. 2052 . “A reasonable probability is a probability sufficient to undermine-confidence in the outcome.” Id. at 694 , 104 S.Ct. 2052 . “[A] criminal defendant alleging prejudice must show ‘that counsel’s errors were so serious as to deprive the defendant of a fair trial, a trial whose result is reliable.’” Lockhart v. Fretwell, 506 U.S. 364, 369 , 113 S.Ct. 838 , 122 L.Ed.2d 180 (1993) (holding that the “prejudice” inquiry under Strickland focuses on whether counsel’s performance renders the result of the trial unreliable or the proceeding fundamentally unfair) (quoting Strickland, 466 U.S. at 687 , 104 S.Ct. 2052 ).

Thomas argues that he was prejudiced in the following ways by counsel’s errors: (1) assuming the 1984 Kern County jury selection process was invalid, he was denied his right to a jury drawn from a fair cross section of the community; (2) he was deprived of the right to have the government establish the validity of its jury selection procedure; and (3) he was deprived of a fair opportunity to argue his claim before the California courts. We reject all of these arguments.

First, we note that even if his jury panel had been perfectly representative of the community, it was highly likely that Thomas would have been tried by an all-white jury given the demographics of Kern County.

Second, we reject Thomas’ claim of prejudice because the evidence, both testimonial and physical, against Thomas was so overwhelming that it is hard to believe that any reasonable juror, black or white, would have voted to acquit him. The following brief summary of the evidence should make this apparent.

After entering the victim’s house at night planning to commit a robbery, Thomas instead forced the victim to perform numerous sex acts. During this time, Thomas asked the victim if she knew someone named “Spider,” to which she replied “no.” After Thomas left the next morning, the victim telephoned two friends who, on their way over to the victim’s house, spotted a man known to them as “Spider.” When the friends arrived at her house, the victim was hysterical. Based upon the description given by the victim, her friends realized that “Spider” was her attacker.

When police arrived at “Spider’s” residence, Thomas’ mother allowed them to enter. The police found clothing matching that described by the victim as the clothing worn by her attacker. Thomas was found crouched in a corner of a bedroom hiding beneath a large quantity of clothing. After healing officers discuss how to package the clothing to preserve it as evidence, Thomas pulled the clothing onto the floor and stomped on it.

Later that morning, Thomas escaped out the back door of the police station, but was promptly reapprehended. Vaginal swabs taken from the victim were consistent with Thomas’ body fluids, and pubic hair identical to Thomas’ was found in the victim’s bed. Moreover, pubic hair matching the victim’s was found in a combing of Thomas’ pubic hair at the time of his arrest. Given this overwhelming evidence of guilt, it is difficult to see how a different jury would have reached a different result.

Finally, we reject Thomas’ argument that he suffered prejudice in his appeals before the state courts. As noted above, in some California cases, courts have employed the comparative disparity test, under which it is easier to meet the second Duren prong than under this court’s absolute disparity test. Thomas’ appellate counsel presented a fair cross-section claim on direct appeal, but the claim was rejected by the state court because there was no evidence to support the claim in the trial record. Again, there is no dispute that the lack of evidence in the trial record was at least partly due to trial counsel’s failure to challenge the jury panel.

While Thomas may have had an easier time making out a prima facie case of under-representation in state court, it is difficult to determine his likelihood of success in meeting the second prong of the Duren test given the absence of the crucial statistics. Nonetheless, even if we assume that Thomas could have satisfied the second Duren prong before a California court, we find it unlikely, based on more recent California cases, that he could have established the systematic exclusion necessary to satisfy the third prong. See Lockhart, 506 U.S. at 366 , 113 S.Ct. 838 (holding that counsel’s failure to make an objection in a state criminal sentencing proceeding, which objection would have been supported by a decision that was subsequently overruled, does not constitute prejudice).

Thomas argues that a possible cause for the alleged underrepresentation of blacks on 1984 Kern County jury panels was the financial hardship excuse policy used in Kern County at that time. According to Thomas, that policy effectively excused low-skilled, non-union employees, whose employers generally do not pay for jury service, and who presumably were disproportionately black. In support of this theory, he cites Alexander, 163 Cal.App.3d at 1199-1203 , 210 Cal.Rptr. 306 , and Buford, 132 Cal.App.3d at 298-99 , 182 Cal.Rptr. 904 .

In Alexander, the court was faced with a challenge to the racial composition of a jury venire in Kern County in August of 1982. Thomas presented evidence in connection with his federal habeas petition that the hardship excuse policy remained the same in Kern County from 1981 through 1984. Nonetheless, Thomas’ reliance on Alexander is misplaced, in that the Alexander court did not find that the excuse policy resulted in systematic exclusion. Rather, it remanded to the trial court for further findings, stating that, “without more factual information, accurate conclusions regarding the cause of the disparity cannot be drawn and any attribution of the cause is wholly speculative.” Alexander, 163 Cal.App.3d at 1201 , 210 Cal. Rptr. 306 . Likewise, the Buford court did not find the challenged excuse procedures in Contra Costa County invalid. Buford, 132 Cal.App.3d at 299 , 182 Cal.Rptr. 904 .

More recently, in People v. Bell, the California Supreme Court held that the defendant failed to make a prima facie showing that a Contra Costa County hardship deferral and excuse policy resulted in systematic exclusion. Bell, 49 Cal.3d at 530 , 262 Cal.Rptr. 1 , 778 P.2d 129 . In so holding, the court noted that such policies are neutral with respect to race, and the court declined to accept the proposition that “constitutionally impermissible ‘systematic exclusion’ could be shown by demonstrating that underrepre-sentation of a cognizable class is the result of applying neutral criteria to grant deferrals on the grounds of hardship.” Id. As noted in Bell , “[n]o case holds ... that disparity that results notwithstanding application of neutral and presumptively constitutionally permissible jury selection criteria is a-product of ‘systematic exclusion.’ ” Id.

Thus, even if Thomas at some point in time could have made out a prima facie case of systematic exclusion based on Kern County’s race-neutral hardship deferral policy, subsequent California case law easts serious doubt on the validity of his claim at the present time. Consequently, Thomas has failed to establish that he was prejudiced before the California courts by trial counsel’s errors. See Lockhart, 506 U.S. at 372 , 113 S.Ct. 838 (holding that prejudice does not result if the ineffectiveness of counsel does not deprive the defendant of any right to which the law entitles him).

CONCLUSION

Thomas has failed to establish a prima facie violation of the Sixth Amendment’s fair cross-section requirement. Thomas has likewise failed to establish his Sixth Amendment claim of ineffective assistance of counsel. Therefore, the district court properly denied his petition for writ of habeas corpus.

AFFIRMED.

. Thomas has abandoned his claim of ineffeclive assistance of appellate counsel for purposes of the present appeal. . Consistent with California law, the term ‘jury panel” is used in this opinion to refer to those persons selected at random from the jury venire (also known as the "juror pool”) and assigned to a particular courtroom for the purpose of selection to try a particular case. See Cal.Code Civ. Proc. § 194(q) (defining “trial jury panel”); see also People v. Bell, 49 Cal.3d 502 , 520 n. 3, 262 Cal.Rptr. 1 , 778 P.2d 129 (1989). . Thomas is not challenging the representation of Hispanics on his jury panel. He merely presents these figures as "circumstantial evidence of systematic exclusion of minorities, including blacks.” . Using these same numbers, the comparative disparity is 100% [ (5%-0%)/5%].