dissenting:
Constantino Carrera, who is Hispanic, was tried and convicted in 1983 of the double murder of a white couple in Kern County, California. In the course of jury selection, the prosecutor exercised the State’s peremptory challenges against six of eight Hispanic venire persons. Although People v. Wheeler, 22 Cal.3d 258, 148 Cal.Rptr. 890, 583 P.2d 748 (1978), had been the law in California for five years, defense counsel did not raise a Wheeler challenge against the prosecution’s highly *952disproportionate use of its peremptory-challenges to excuse 75 percent of Hispanic venire persons who were called. Because, in failing to object and failing to develop the record, Carrera’s trial counsel ignored an obvious prima facie ease that the prosecutor was using his peremptory challenges disproportionately to strike Hispanic jurors in violation of Wheeler, I respectfully dissent from the majority’s denial of the writ.1 Because this deficient performance resulted in a structural error, prejudice must be presumed. Accordingly, I would grant Carrera habeas relief.
I
Carrera has overcome the “strong presumption” that his trial counsel’s failure to object was “sound trial strategy.” Strickland v. Washington, 466 U.S. 668, 689, 104 S.Ct. 2052, 80 L.Ed.2d 674 (1984). Trial counsel could not offer any reason or explanation for her failure to make a Wheeler challenge. In these circumstances, Carr-era has established that his trial counsel failed properly to develop the record and to object where there was an obvious prima facie case of the biased use of peremptory challenges. This was not “sound trial strategy” — insofar as the record shows, it was no strategy at all — nor does it fall within the “wide range of reasonable professional assistance.” Id.
Under Wheeler, a party may establish a prima facie case of the discriminatory use of peremptory challenges by “showing] that his opponent has struck most or all of the members of the identified group from the venire, or has used a disproportionate number of his peremptories against the group.” 148 Cal.Rptr. 890, 583 P.2d at 764. This statistical evidence is bolstered if, as is the case here: (1) the defendant is a member of the excluded group; and (2) the victim is a member of the group to which the majority of remaining jurors belong. Id.
Had she objected, Carrera’s trial counsel clearly could have made a prima facie case that the prosecutor used his peremptory challenges to eliminate “members of a cognizable group”: Hispanics. The prosecution used its peremptory challenges to strike 75 percent of potential Hispanic jurors and only 27 percent of the potential non-Hispanic white jurors. Carrera is Hispanic. The victims were white. This is a prima facie case of bias under Wheeler.
Was the fairness of Carrera’s trial compromised because the prosecutor used his peremptory challenges to strike most of the Hispanic potential jurors on the basis of group bias? Of course, we will never know the answer. Because Carrera’s trial *953counsel failed to object, the prosecutor never had to justify his peremptory challenges.2 Indeed, Carrera’s counsel failed even to make a record, so there is little to indicate what the prosecutor’s “group bias”-neutral reasons may have been, if any, for striking the potential Hispanic jurors.3 This is not the sort of small, inevitable trial error which led the Supreme Court to observe, “A defendant is entitled to a fair trial but not a perfect one.” Lutwak v. United States, 344 U.S. 604, 619, 73 S.Ct. 481, 97 L.Ed. 593 (1953). Rather, it is the kind of fundamental error that calls into question the basic fairness of the trial mechanism itself. It cannot, and should not, be explained away by speculation on what possible reasons might have motivated either the prosecutor or defense counsel.
In the absence of any record evidence to support its position, the majority searches for snippets in the voir dire transcript to justify that which, prima facie, appears motivated by bias. “But it does not matter that the prosecutor might have had good reasons to strike the prospective jurors. What matters is the real reason they were stricken.” Paulino v. Castro, 371 F.3d 1083, 1090 (9th Cir.2004).4 The majority engages in exactly the kind of “judicial speculation” of which the Supreme Court expressly disapproves. See Johnson v. California, 545 U.S. 162, 173, 125 S.Ct. 2410, 162 L.Ed.2d 129 (2005) (criticizing “the imprecision of relying on judicial speculation to resolve plausible claims of discrimination”). The majority fails to explain why it is appropriate, on habeas review, to do exactly what trial courts have been instructed they absolutely may not do: speculate about the prosecutor’s reasons for striking potential jurors who belong to a cognizable group.5
Critically, Wheeler does not require the removal of a specific number or percentage of the targeted group in order to establish a prima facie violation. See 148 Cal.Rptr. 890, 583 P.2d at 764 (explaining that a party can make a prima facie case by “showing] that his opponent has struck most or all of the members of the identified group from the venire”) (emphasis added). A violation occurs, and a new jury must be drawn, if even a single peremptory was based on group-bias. See id., 148 Cal.Rptr. 890, 583 P.2d at 765. As the California Supreme Court stated:
If the court finds that the burden of justification is not sustained as to any of the questioned peremptory challenges, the presumption of their validity is re*954butted. Accordingly, the court must then conclude that the jury as constituted fails to comply with the representative cross-section requirement, and it must dismiss the jurors thus far selected. So too it must quash any remaining venire, since the complaining party is entitled to a random draw from an entire venire — not one that has been partially or totally stripped of members of a cognizable group by the improper use of peremptory challenges. Upon such dismissal a different venire shall be drawn and the jury selection process may begin anew.
Id. (emphases added). Accordingly, the absence of a “group bias”-neutral reason for striking even one potential juror constitutes a Wheeler violation.
The record does not support the majority’s conclusion that there were reasons neutral of “group bias” to strike all six of the Hispanic potential jurors, and that such reasons were so obvious or apparent as to excuse trial counsel from making a Wheeler motion. The majority admits that, even under its speculative review of the record, there was no obvious reason for striking potential juror Martinez. When one compares the voir dire responses of Martinez (who was struck) and Allen (a non-Hispanic white juror who was not struck), no obvious, permissible reason emerges for striking one and not the other, both of whom lived at least thirty miles from the courthouse.
The majority speculates that Martinez was excused because he was disabled, even while admitting that the record does not disclose what his disability was. Maj. Op. at 947-48. The majority so speculates in spite of the fact that Martinez was qualified to serve.6 See Maj. Op. at 941 & n. 4.
Indeed, a closer look at the voir dire of Allen reveals the folly of the majority’s approach of parsing the record to construct hypothetical reasons for dismissing or not dismissing potential jurors. Asked whether he was familiar with the case, Allen stated that he “heard it in the newspapers” and he had co-workers who lived at the Imperial Motel where the murders took place. If Mr. Allen had instead been “Mr. Allende,” and subject to a peremptory challenge by the prosecution, the majority would have concluded that the dismissal was “group bias”-neutral because the potential juror was familiar with the case and the scene of the crime.
Importantly, trial counsel’s failure to make a Wheeler motion was not based on trial strategy. See Strickland, 466 U.S. at 689, 104 S.Ct. 2052. During state habeas proceedings, Carrera’s trial counsel testified that she did not know why she failed to make a Wheeler motion during voir dire. Indeed, she has never articulated a strategic reason for failing to make a Wheeler motion.
Of course, trial counsel’s utter failure to present a reasonable explanation for her failure to object would be of no moment if it fell “within the range of reasonable representation.” See Morris v. California, 966 F.2d 448, 456 (9th Cir.1991) (relying on defense counsel’s explanation that his decision not to call a certain witness “was a tactical one,” made to avoid linking the defendant “to prior involvement with cocaine”). It does not. In spite of the majority’s hypothetical musings, no profes*955sional strategic decision can be discerned for trial counsel’s failure to object to the disproportionate striking of Hispanic , potential jurors. It is telling that the State does not even attempt to find “group bias”-neutral reasons in the record, and the majority’s concluding speculation is only that Carrera’s “counsel may have been pleased with the resulting jury.” Maj. Op. at 949. See also Maj. Op. at 950 (stating the “possibility that defense counsel was pleased with the resulting jury”).
Given that the defendant was Hispanic and the victims were white, Carrera’s lawyer should have picked up on the prosecutor’s apparent attempt to skew the jury. It does not take an expert to conclude that something is fishy — it simply does not pass the smell test — when a prosecutor strikes 75 percent of Hispanics on a jury and only 27 percent of non-Hispanie whites. The majority speculates that “defense counsel may have been pleased with the resulting jury, despite the fact that the prosecutor had removed several Hispanic venirepersons.” Maj. Op. at 949. She may have been “pleased” (although the majority does not say why she, or any reasonable attorney, would have been), but that is not the question. The question is whether a reasonable attorney in the same circumstances would have objected. Any reasonable attorney would have done so. At the very least, a reasonable attorney would have developed the record. Accordingly, the failure to bring a Wheeler motion constituted deficient performance.7
II
We have held that the discriminatory use of peremptory challenges, known in *956federal law as Batson error,8 is a structural error, rather than trial error, that requires reversal. E.g., Kesser v. Cambra, 465 F.3d 351 (9th Cir.2006) (en banc). The discriminatory use of peremptory challenges is not merely a trial error which can be weighed in the context of other admitted evidence; it is a defect in the trial mechanism. Thus, I would hold that Wheeler error, like Batson error, is structural error, and prejudice must be presumed. See Arizona v. Fulminante, 499 U.S. 279, 309-10, 111 S.Ct. 1246, 113 L.Ed.2d 302 (1991) (explaining that unlike a trial error, structural errors require reversal because they “affect[] the framework within which the trial proceeds, rather than simply an error in the trial process itself’).
Where ineffective assistance of counsel results in a structural error, prejudice must be presumed. See Owens v. United States, 483 F.3d 48, 64 n. 14 (1st Cir.2007) (presuming prejudice where ineffective assistance of counsel led to a closed trial, explaining that “prejudice is presumed in cases of structural error not because the risk of prejudice is high, but because it is impossible to determine the extent of the prejudice”); see also McGurk v. Stenberg, 163 F.3d 470, 475 (8th Cir.1998) (“[W]hen counsel’s deficient performance causes a structural error, we will presume prejudice under Strickland.”); Gov’t of Virgin Islands v. Forte, 865 F.2d 59, 63-64 (3d Cir.1989). These cases recognize that structural error exists because, in certain cases, prejudice is impossible to quantify. It hardly makes sense to require a petitioner “to do what the Supreme Court has said is impossible.” Owens, 483 F.3d at 65.
This is particularly true with respect to a Wheeler or Batson error. See Powers v. Ohio, 499 U.S. 400, 412, 111 S.Ct. 1364, 113 L.Ed.2d 411 (1991) (“Active discrimination by a prosecutor during [jury selection] condones violations of the United States Constitution within the very institution entrusted with its enforcement, and so invites cynicism respecting the jury’s neutrality and its obligation to adhere to the law.”). The Supreme Court further stated:
A prosecutor’s wrongful exclusion of a juror by a race-based peremptory challenge is a constitutional violation committed in open court at the outset of the proceedings. The overt wrong, often apparent to the entire jury panel, casts doubt over the obligation of the parties, the jury, and indeed the court to adhere to the law throughout the trial of the cause. The voir dire phase of the trial represents the jurors’ first introduction to the substantive factual and legal issues in a case. The influence of the voir dire process may persist through the whole course of the trial proceedings.
Id. (internal quotation marks and citation omitted).
To show actual prejudice, the petitioner would be required to put forth evidence as to whether there was a reasonable probability that a jury with a different racial composition would have reached a different verdict. See Strickland, 466 U.S. at 694, 104 S.Ct. 2052. But the “consequences [of a structural error] ... are necessarily unquantifiable and indeterminate,” which would place an impossible burden on the petitioner. Sullivan v. Louisiana, 508 U.S. 275, 282, 113 S.Ct. 2078, 124 L.Ed.2d 182 (1993). Requiring a petitioner to prove prejudice in such a situation would prevent structural errors, the most serious type of error, from forming the basis of an ineffective assistance of counsel claim. Thus, I would hold that *957prejudice must be presumed from defense counsel’s failure to make a Wheeler motion in response to the State’s discriminatory use of peremptory challenges.
III
Because I would grant a conditional writ of habeas corpus requiring a new trial with a Wheeler (and Batson) compliant jury, I respectfully dissent.
. Although the district court conducted a careful review of the voir dire of the eight Hispanic jurors, as the majority recognizes, Maj. Op. at 941 n. 3, the district court erred in its Wheeler analysis. The district court concluded that Carrera failed to show that there was a "strong likelihood of discriminatory purpose,” citing two California Court of Appeal cases which held that a Wheeler motion could not be successful if the prosecutor left two or three members of the cognizable group on the jury. See Carrera v. Ayers, 2008 WL 681842, at *26 (E.D.Cal.2008) (citing People v. Davis, 189 Cal.App.3d 1177, 234 Cal.Rptr. 859 (1987); People v. Boyd, 167 Cal.App.3d 36, 212 Cal.Rptr. 873 (1985)). But these cases were not decided until two and four years after Carrera’s trial in 1983, so they could not have been the law in California at the time of Carrera’s trial. Moreover, these cases were quickly overruled by the California Supreme Court. See People v. Snow, 44 Cal.3d 216, 242 Cal.Rptr. 477, 746 P.2d 452 (1987). Thus, although the district court acknowledged that, under Strickland, it was required to assess Carrera’s trial counsel's "performance in the context of what the applicable law was at the time of Carrera’s trial,” it nevertheless erred by evaluating counsel's effectiveness on the basis of case law that was not the law of California at the time of Carrera's trial.
. Tellingly, when asked in post-trial proceedings, the prosecutor could not recall his reasons for striking five of the six potential Hispanic jurors, although he denied that his actions were racially motivated. This gives rise to a reasonable inference that there was no apparent reason, other than group bias, to challenge them.
. Trial counsel’s failure even to make a record on the dismissal of the potential Hispanic jurors, particularly potential juror Martinez, was deficient performance. See Wheeler, 148 Cal.Rptr. 890, 583 P.2d at 764 (explaining that a party who suspects his opponent is striking potential jurors based on group bias ''should make as complete a record of the circumstances as is feasible”).
. As the majority notes, Maj. Op. at 941-42, this is a pre-AEDPA case.
. This differs from the legitimate use of comparative juror analysis, which takes place at the third step of a Wheeler or Batson challenge, after the prosecution proffers bias-neutral reasons for striking members of a cognizable group. See Lewis v. Lewis, 321 F.3d 824, 830-31 (9th Cir.2003) (explaining that the court may review the voir dire record to determine whether the prosecutor’s proffered reasons are credible or pretextual); Wheeler, 148 Cal.Rptr. 890, 583 P.2d at 765.
. There is no end to the majority’s speculation. Its own speculation would have the prosecutor use his peremptory strikes in a manner contrary to law: ”[T]he prosecutor wanted to ingratiate himself with the remaining jurors by relieving the disabled Martinez from sitting on a long jury trial.” Maj. Op. at 947. But it has been well-established for many years that the law forbids discriminating against the disabled in jury service. See, e.g., Greater L.A. Council on Deafness, Inc. v. Zolin, 812 F.2d 1103 (9th Cir.1987).
. The majority cites three completely inapposite cases in an attempt to bolster its conclusion that Carrera’s trial counsel’s performance was not deficient. None of these cases deals with a defense attorney’s failure to monitor the prosecutor’s voir dire for signs of group bias. Rather, each of the cases cited by the majority deals with ineffective assistance of counsel claims premised on the defense attorney’s failure to challenge a single, possibly biased juror or to ask more questions during voir dire. First, the majority relies on United States v. Quintero-Barraza, 78 F.3d 1344 (9th Cir.1995), for the proposition that this court defers to counsel’s decisions during jury selection. Maj. Op. at 948-49. In fact, we held that we must be "highly deferential” where trial counsel makes a tactical decision in jury selection whether to challenge a potential juror on the ground of bias. 78 F.3d at 1349. We therefore rejected Quintero-Barraza’s ineffective assistance of counsel claim premised on defense counsel’s failure to strike a single juror where counsel explained that he considered the juror to be honest and impressive, notwithstanding the juror’s statement that he would find it difficult to be impartial. Id. Second, the majority cites Fields v. Woodford, 309 F.3d 1095 (9th Cir.2002). Maj. Op. at 948-49. Fields was decided on the ground of absence of prejudice. Id. at 1107-08 (remanding for findings on the possible bias of a single juror but declining to conclude that the defendant's counsel was ineffective for failing to question other jurors who did not appear possibly biased on the record, and so could not have prejudiced the result). Finally, the majority relies on Hovey v. Ayers, 458 F.3d 892 (9th Cir.2006), see Maj. Op. at 949, in which we held that defense counsel was not deficient where his limited questioning during voir dire was reasonably explained as a tactical decision to not remind the jurors of publicity the case had received four years before the trial. Id. at 910.
Here, Carrera does not question how his trial counsel exercised her own peremptory challenges or her questioning of potential jurors. Instead, he challenges his trial counsel’s utter failure to object to the facially discriminatory manner in which the prosecutor exercised his peremptory challenges. Most importantly, unlike the cases relied on by the majority, there is no evidence in the record that Carrera’s trial counsel had any tactical reason for her failure to act in the face of a prima facie case of group bias. We cannot be “highly deferential” to a tactical decision that does not exist in the record and, indeed, one trial counsel herself admits she never made.
. See Batson v. Kentucky, 476 U.S. 79, 106 S.Ct. 1712, 90 L.Ed.2d 69 (1986).