concurring in part:
I concur fully in Judge Clifton’s analysis of Gonzales’s non -Brady claims.
I also concur in the decision to remand to the district court with directions to stay and abey Gonzales’s federal habeas petition in order to allow him to present to the California state courts his claim under Brady v. Maryland, 373 U.S. 83, 83 S.Ct. 1194, 10 L.Ed.2d 215 (1963), based on the evidence he has obtained during discovery in his federal habeas proceedings. However, I believe that we have the authority, in the circumstances of this case, consistent with the Supreme Court’s decision in Cullen v. Pinholster, — U.S.-, 131 S.Ct. 1388, 179 L.Ed.2d 557 (2011), to decide petitioner’s Brady claim now, without first seeking the view of the California courts. I would prefer to exercise that authority.
I. General Application of Pinholster
The Supreme Court held in Pinholster that in a habeas case governed by 28 U.S.C. § 2254(d)(1) we generally cannot *1000consider evidence that was not before the state court. Pinholster, 131 S.Ct. at 1398. A question Pinholster does not answer is what federal courts should do when confronted with such evidence. I agree with Judge Clifton that we may treat such evidence in the same way we treat unexhausted claims.
In Rhines v. Weber, 544 U.S. 269, 125 S.Ct. 1528, 161 L.Ed.2d 440 (2005), the Supreme Court held that when a habeas petitioner files a so-called “mixed” habeas petition — a petition that includes both exhausted and unexhausted claims — a district court abuses its discretion in certain circumstances when it decides not to stay and abey the petition to allow the petitioner to present the unexhausted claims in state court. Specifically, a district court should stay and abey such a mixed petition if “the petitioner had good cause for his failure to exhaust, his unexhausted claims are potentially meritorious, and there is no indication that the petitioner engaged in intentionally dilatory litigation tactics.” Id. at 278, 125 S.Ct. 1528. The Court reasoned that if district courts did not stay and abey such mixed petitions, the one-year statute of limitations in AEDPA would bar petitioners from obtaining federal review of the unexhausted claims. See id. at 275, 125 S.Ct. 1528. Restricting the stay and abeyance procedure to situations where petitioners have good cause for failure to exhaust, have potentially meritorious claims, and have not engaged in dilatory tactics, protects “AEDPA’s objective of encouraging finality.” Id. at 277, 125 S.Ct. 1528.
The same reasoning applies here. If a habeas petitioner discovers new evidence supporting a claim of a constitutional violation after he has filed a federal habeas petition, and the district court adjudicates that claim without the benefit of the new evidence, then the ban on second or successive petitions in 28 U.S.C. § 2244 will usually bar the petitioner from ever having his full claim adjudicated in federal court. I agree with Judge Clifton that, so long as the petitioner had good cause for his failure to discover the evidence while before the state court, the claim based on the new evidence is potentially meritorious, and there is no indication that the petitioner engaged in intentionally dilatory litigation tactics, the district court should stay and abey the claim supported by the new evidence to allow the petitioner to present that evidence to the state court. I also agree with the application of Rhines to the facts of this case.
If this were an application of Pinholster to an ordinary case, I would simply agree with Judge Clifton’s opinion and not write separately.
II. General Rule of Pinholster Does Not Apply
This, however, is not an ordinary case. The facts of this case are so extreme that the federal courts have the power, consistent with Pinholster, to consider the Brady material now, without first requiring Gonzales to seek the view of the state courts.
The prosecutors actively suppressed evidence concerning William Acker, the witness who gave the only evidence of the state’s primary aggravating factor at Gonzales’s capital trial. The California trial court on post-conviction review granted a discovery motion that would have revealed the suppressed evidence. But the California Supreme Court reversed the trial court’s discovery order and refused to enter one of its own. The suppressed evidence shows not only that Acker was a jailhouse informant expecting to receive advantageous treatment in return for his testimony. Acker was also, unbeknownst to Gonzales and his attorneys, a schizo*1001phrenic who had repeatedly faked suicide attempts to obtain desired prison transfers; had attempted to bribe a prison psychologist for a good report; had previously said that he was turning his life around in order to obtain release from prison, but had committed thefts and a first-degree murder upon his release, before testifying at Gonzales’s trial that he was testifying as part of an effort to turn his life around; and had earlier been described by state correctional officers as “considered capable of any measure of brutality in the service of achieving what he wants to do.”
I explain below the procedural history of Gonzales’s Brady claim, and why, given that history, Pinholster does not prevent us from deciding that claim now.
A. Acker’s Role at Trial
Judge Clifton accurately describes the events of Gonzales’s trial. I provide more detail to show fully Acker’s importance to the state’s case against Gonzales, and the extent to which Acker was — or, more accurately, was not — impeached at trial.
1. Acker’s Substantive Testimony
Gonzales’s conviction was based on his murder of a police officer. The officer was part of a team conducting a drug raid of Gonzales’s parents’ home. When the police entered, Gonzales was braced against a wall pointing a shotgun at the door. He fired and killed one of the officers. Gonzales’s guilt-phase defense was that he thought the people entering his house were not police, but rather were members of the rival Bassett gang.
The prosecution’s evidence supported two somewhat different stories. The first was that Gonzales must have known, based on the facts on the day of the raid, that the people breaking open his parents’ front door were police officers. As Judge Clifton describes, the police officers testified that they repeatedly announced they were police before entering and that they had entered the house with their badges visible. Another officer with expertise in gang activity testified that the area where Gonzales’s parents lived did not have a high level of gang activity, that the gang activity that did take place consisted of drive-by shootings rather than home invasions, and that a typical member of the Bassett gang looked and dressed nothing like the plain-clothes officers who conducted the raid. This evidence suggested it was unlikely that Gonzales mistook the plain-clothes officers for members of the Bassett gang.
The second story was that Gonzales had been told in advance that the police were coming to conduct the raid, and that he had formed, well in advance of the raid, an intent to kill a police officer. This story, obviously much more damning to Gonzales, came entirely from the testimony of William Acker. Acker was a jailhouse informant. Acker’s story was generally consistent through Gonzales’s guilt trial, first penalty-phase trial, and second penalty-phase trial. Acker testified that Gonzales had approached him while they were in prison cells in the same row shortly after Gonzales was arrested. Acker testified that in discussions during the next two weeks, Gonzales told him the details of his crime. According to Acker, Gonzales told him that he had been tipped off by a friend that the police were going to raid his house. Gonzales thus knew the police were coming and was “on point” all day waiting, in Acker’s words, to “bag a cop.” According to Acker, Gonzales “knew it was a cop. He knew he was going to kill one. He aimed to kill.” Acker testified that when he asked Gonzales why he was so determined to kill a police officer, Gonzales described to him “a real sick philosophy about how to protect the pad.” Gonzales *1002wanted to shoot a cop because “[t]hey got that coming.” Acker also testified that Gonzales had told him that his defense strategy would be to claim that he thought the police officers were members of the Bassett gang.
2. Acker’s Importance
Acker’s testimony was important in the guilt phase because, if believed, it demonstrated that Gonzales knew well in advance of the raid that the men conducting the raid were police officers. However, all the prosecution needed to show in the guilt phase was that Gonzales knew that the men were police officers at the time he fired. Acker’s testimony was thus useful but not essential to the guilty verdict. In the penalty phase, however, Acker provided the crucial testimony. In arguing why Gonzales deserved the death penalty, the prosecution’s main theory of aggravation was that Gonzales had planned in advance to “bag a cop.” In his opening statement in the second penalty-phase trial, Ronald Bowers, the prosecutor, said that “the evidence will show that the defendant planned to kill a cop. Not that it was thought of lightly, that it was just a split-second situation, spontaneous, but he planned to kill a cop. Something to the effect he was going to bag a cop.” The evidence that Gonzales had planned in advance to kill a police officer came only from Acker, and the expression “bag a cop” came only from Acker. In his closing statement, Bowers’s chronology of when and how Gonzales formulated his plan to “bag a cop” came entirely from Acker’s testimony.
Acker’s importance in the penalty phase is also demonstrated by a memorandum Bowers wrote to a colleague after Gonzales’s first penalty-phase trial resulted in a mistrial. The jury, which consisted of six men and six women, had hung with nine in favor of death and three in favor of life without parole. Bowers wrote:
I talked to the jurors afterwards and most of them found William Acker’s testimony to be credible. It was interesting that the women (especially the younger women) were totally convinced by Acker’s statements. Some of the men only partially believed what he said. I hope to retry the penalty phase if downtown approves the retrial and Bill Acker is available and is willing to testify.
This memorandum makes clear that Bowers was not willing to retry the penalty phase without Acker. It further suggests that women were more likely to believe Acker. The jury in the second penalty-phase trial had eleven women and one man. Gonzales’s trial took place before the Supreme Court decided J.E.B. v. Alabama, 511 U.S. 127, 114 S.Ct. 1419, 128 L.Ed.2d 89 (1994), and it was thus permissible for the prosecutor to seek women on the jury.
Attacking Acker’s truthfulness was crucial to the penalty-phase defense strategy. Ralph Bencangey, Gonzales’s lawyer, decided not to attack the guilt-phase jury’s conclusion that at the time he fired the gun Gonzales knew the men were police officers, but to rather attack Acker’s testimony that Gonzales had planned in advance the murder of a police officer. Bencangey told the second penalty-phase jury that the guilt-phase jury “could well have thought that at the time it occurred and the way that it occurred, [Gonzales] should have known that they were police officers at that time rather than ... a long period in advance.” Bencangey told the jury in his opening statement that he would show that “if [Gonzales] is guilty, as the jury has previously stated, it would not be based on some advance plan that the prosecution is talking about, some special plan to kill a *1003police officer.” Bencangey made this same argument in his closing.
To succeed with this strategy, it was crucial to impeach Acker. In his opening statement Bencangey said that “the People are asking for a death penalty verdict, really based on the absolute truth of everything that Mr. Acker has said. It must be absolutely true.”
3. Impeachment of Acker
Over the course of one guilt-phase trial and two penalty-phase trials, the prosecution disclosed only two documents concerning Acker to the defense: Acker’s rap sheet and a transcript of Acker’s initial statement to investigators. The rap sheet was incomplete and inaccurate. It included several property crimes and Acker’s conviction for first-degree murder. But it failed to include several of Acker’s convictions, and it incorrectly stated that Acker was sentenced to life without parole for the murder when he was actually sentenced to life with possibility of parole. Acker’s initial statement to investigators provided no impeachment material because Acker told them basically the same story he later told at trial.
Bencangey was thus severely limited in his ability to impeach Acker. During cross-examination in the guilt-phase trial, Acker admitted he was hopeful that in exchange for his testimony he would be transferred to an out-of-state prison. He said he was interested in such a transfer because he hated gangs, and the gangs hated him in return. Acker also admitted that he had given information to the police about crimes on two prior occasions. First, he gave information about his wife’s involvement in a murder that he and his wife had committed in Hawaii. Second, he spoke with the police about a murder that had taken place on his row in jail. Everyone on his row, however, had been required to speak to the police about that murder.
By the time of the first penalty-phase trial, Acker had given information in several more cases. In addition to having given information about his wife and the jailhouse murder, Acker admitted giving information to the police in the cases of four other criminal defendants: Anthony Davis, Raymond LaScola, Edwin Guy Williams, and Johnny Torres. In violation of Brady, the prosecutor had not disclosed to Gonzales Acker’s involvement in these cases. Bencangey had nonetheless learned of Acker’s involvement when he discovered the transcript of Acker’s testimony in the preliminary hearing in the LaScola case. In all four of the cases, Acker claimed the defendant confessed to him. In some of them, Acker testified in court, and in some he only spoke to the police. Although Bencangey did not know it during trial, we now know that the LaScola case was dropped because then-Head Deputy District Attorney (now our colleague) Stephen Trott had concluded that Acker was an unreliable witness. Judge Trott’s view, expressed in a deposition taken in connection with Gonzales’s federal habeas petition, was that Acker’s testimony was “certainly nothing that a jury ever should rely on if that’s all there was to convict Dr. LaScola.” Judge Trott described Acker as a “classic psychopath” and “a truly evil guy who delighted in sticking guns in people’s mouths to try to get them to, quote, ‘crap in their pants.’ ”
To justify his practice of informing, Acker testified that not only was he giving information in order to obtain a transfer to an out-of-state prison, but he was also seeking a new “moral balance.” He testified that he wanted to “do something right for a change, to do something real.” “I’m trying to balance out the wrong that I’ve been through, you know, stop somebody *1004else from going out and killing someone again.”
In the second penalty-phase trial, Acker again discussed the information he had provided in the other cases, and he discussed his involvement in yet another case. He again testified that his reason for providing this information was both his desire to transfer to an out-of-state prison and his desire to “start a different direction for myself.” Before the second penalty-phase trial, Bencangey had discovered that the state had promised Acker that they would remove a tattoo from his back that read “Paramount.” When Bencangey attempted to question Acker about the tattoo, Bowers objected. In a conference with the judge, Bowers said he wanted to prevent questioning about Acker’s tattoo because “we have tried to change the identity of this informant, for his own safety.” Bencangey replied that “up until this hearing, until I found out, I’d asked that witness if anything else was offered to him in exchange for his testimony, and the district attorney’s office has never been forthcoming in telling me that they had promised to change his identity and surgically remove tattoos off him and give him a different name. I’ve only found this out recently.”
The trial judge allowed Bencangey to question Acker about the tattoo. Bencangey attempted to impeach Acker by suggesting that the tattoo was a symbol of Acker’s gang membership, and that the tattoo removal and the identity change was another part of what the state had offered him in exchange for testifying. Acker denied that he was in a gang, claiming that he had the tattoo to represent that he was from the city of Paramount.
The second penalty-phase jury, consisting of eleven women and one man, returned a unanimous verdict in favor of death.
B. Report of the 1989-90 Los Angeles County Grand Jury
Gonzales was sentenced to death in May 1981. In October 1988, Los Angeles Sheriffs Department deputies learned that a jailhouse informant, Leslie White, was writing an article in which he planned to explain how informants were acquiring knowledge about defendants they had never met in order to fabricate those defendants’ confessions. Report of the 1989-90 Los Angeles County Grand Jury, June 26, 1990, at 69 [hereinafter “Grand Jury Report”]. White agreed to demonstrate the technique to the Sheriffs Department. The report of what White did is worth repeating in full. The report does not identify White by name, but refers to him as “the informant.”
The sergeant gave the informant the name of an inmate who was being held in the Hall of Justice Jail on murder charges. The informant, representing himself to be an employee of a bail bond company, called the jail’s Inmate Reception Center and was able to obtain the inmate’s booking number, date of birth, color of eyes and hair, height, weight, race (Caucasian), bail ($100,000), case number, date of arrest, arresting agency (Sheriffs Special Enforcement Bureau), next court date, and where the inmate was housed in the jail.
The informant next called the records section of the District Attorney’s Office. He said he was a Deputy District Attorney and asked for information on the inmate’s case. He was given the name of the Deputy District Attorney prosecuting the case, the Deputy District Attorney’s telephone number, and the name of a witness.
A few calls later, the informant called Sheriffs Homicide and said he was “Sergeant Stevens” at the Central Jail. He *1005was able to obtain the name of the murder victim, and the victim’s age and race.
The informant then called the Deputy District Attorney who was handling the case, initially identifying himself as “Sergeant Williams” with the Los Angeles Police Department. The Deputy District Attorney responded to the informant’s questions by stating, “I’ll tell you anything you want to know about the case,” and proceeded to provide details about what the victim was wearing, where his body was found, the fact that the coroner’s report said that death resulted from suffocation and/or drugs, that the victim’s blood contained a fatally high amount of methamphetamine, that the defendant confessed to stuffing the victim in a trunk, and the prosecutor’s personal opinion of the likely defense in the case. Near the end of the conversation, the informant gave his name as “Sergeant Johnson.”
At this point, the informant said he had obtained enough details about the case to enable him to fabricate a jail house confession which would be accepted by detectives. He then proceeded to demonstrate how he could arrange for contact between himself and the inmate to support the fabricated confession.
The informant called a department of the Superior Court in Van Nuys, identifying himself as Deputy District Attorney “Michaels” with the Organized Crime Unit downtown. In response to the informant’s request, the court bailiff ordered the informant and the inmate to be transported to Van Nuys the following day.
Id. at 69-71. White indicated that a fabricated confession based on this information would be sufficient to get him a “hell of a deal” with prosecutors. Id. at 71. His information was believable both because he knew many facts about the defendant that he apparently could only have obtained from the defendant, and because he had arranged to be in the same place as the defendant so that it was possible that the defendant could have confessed to him. White said that the “key” to the operation is that the District Attorneys
want to win. So if I come forward with the information as detailed as that they’re gonna use it. Because the jury not knowing the system or how it works, is going to believe when I get up there with all these details and facts, that this guy sat in the jail cell, or he sat on the bus, or he sat in the holding tank somewhere, or told me through a door or something, they’re gonna believe me.
Id. at 72.
White’s demonstration led to the empanelling of an investigatory Grand Jury to look into the improper use of informant testimony by the Los Angeles County District Attorney’s Office. The Grand Jury’s investigation took place between 1989 and 1990, and covered the period from 1979 through 1990. The grand jurors and their investigators spoke to twenty-five informants. The Grand Jury concluded that “the experiences and perceptions of these informants generally reflect those of the informant population at large.” Id. at 8. The grand jurors also spoke to and heard testimony from a large number of people in the Los Angeles County Sheriffs Department (which administers the jails where the informants were housed), the Los Angeles District Attorney’s Office, and the defense bar. The Grand Jury’s 150-page report paints a harrowing picture of the role of jailhouse informants in the Los Angeles County criminal justice system during this period.
The report gave a detailed description of jailhouse informants’ “astonishing ability to discover information about crime in or*1006der to concoct a confession by another inmate.” Id. at 31. Credibility of jailhouse informants is often based on the supposed fact that they could only have obtained certain information about a crime from the defendant himself. The report describes cases in which law enforcement officers either fed information about other inmates to informants, or left informants in a room with documents from other inmates’ cases. Id. at 27-28. Sometimes law enforcement officials would signify to informants that a new inmate was “hot,” thereby giving the informants an implicit instruction to provide testimony inculpating the inmate. Informants would sometimes obtain information by getting friends who were not incarcerated to go to a defendant’s preliminary hearing and get information, or would simply get unwitting defendants to describe crimes. In one case, a defendant described to an informant a crime he had witnessed, but not committed, and the informant turned it into a confession. Id. at 30. Often informants would “offer to assist [a] defendant in his case and thereby elicit the defendant’s knowledge of law enforcement’s version of the crime.” Id. at 31 n. 16.
The Grand Jury found that the Sheriffs Department facilitated informants’ collection of information by placing informants near defendants from whom it wanted a confession. According to the Grand Jury:
It has long been suspected that Sheriffs Department deputies intentionally placed informants with inmates “from whom law enforcement could use a confession.” The Sheriffs Department denies such a practice has ever existed, however, the Grand Jury received evidence which indicated the placing of inmates for the purpose of gathering information has occurred.
Id. at 58 (quoting letter of Nov. 1, 1988, from a supervisory Deputy District Attorney to the District Attorney’s Director of Bureau, Branch and Area Operations). The Grand Jury went on to describe numerous instances in which it concluded that informants had been placed near defendants from whom authorities wanted confessions. Id. at 60-68.
In addition to their ability to discover information about defendants, the informants had no scruples about perjuring themselves.
An appalling number of instances of perjury or other falsifications to law enforcement during the past ten years were described by informants. Undeniably, a significant number of informants do not tend to feel constrained by external or internal values to refrain from lying, regardless of the consequences to other inmates.
Id. at 18-19. One informant confessed in the mid-1970s to a crime he had not committed. In 1979, a psychiatrist diagnosed this informant as a pathological liar. Prosecutors subsequently used this informant’s testimony against defendants in either five or six cases. Id. at 16.
The number of informants was staggering. One defense attorney described a case in which, at the preliminary hearing, eight jailhouse informants claimed his client made incriminating statements. Other informants called the attorney and offered to testify on his client’s behalf. The attorney described the preliminary hearing as “a dream world where everybody was either lying or fabricating or then recanting a prior lie or then making something up. It got to be a nightmare where you couldn’t believe one person or the other.” Id. at 39.
High profile cases attracted the most informants. In some cases, up to twenty informants would come forward claiming to have heard a confession. Id. at 111. The Grand Jury found a general consensus *1007among defense attorneys that “jail house informants seemed to invariably be available to testify in important cases.” Id. at 37. Of all the death penalty cases tried in Los Angeles from 1978 through 1990, approximately one third involved a jailhouse informant testifying that the defendant had confessed to him. Id.
The Grand Jury found that informants were given numerous benefits. Examples include being transferred to a cell with a TV, coffeepot, and other amenities, being taken outside the jail for lunch, having witness protection money paid to an informant’s wife, and having a girlfriend being held on a one million dollar bail released on her own recognizance. Id. at 13-15. At least two informants testified to instances where they or an informant they knew were transferred to a jail perceived to be more desirable. Id. at 14. The Grand Jury also found many examples of prosecutors dropping charges, writing letters on informants’ behalf to the parole board, or asking a court to impose a lower sentence. Id. at 13.
Benefits to informants were not generally disclosed to the defense. The prosecutors’ rationale for failing to disclose the benefits was that there was only an implicit understanding that the authorities would help the informant in exchange for his testimony. The Grand Jury explained:
The entire circumstances regarding benefits and the expectations of benefits, in many cases, are not adequately presented to the judge or jury for them to have the necessary factual basis to evaluate the testimony of the informant. This is particularly so when an agreement on the extent of benefits is not made with the informant until after the testimony.
Id. at 76. Later, along similar lines, the Grand Jury found:
The practice of waiting until after the testimony is provided, before the informant’s pending case is dealt with, may lend itself to some troubling results. This may provide the informant with a basis for assuming that his sentence will be measured by the assistance he provides the prosecution by his testimony. In view of the benefits that he may be seeking by his testimony, the potential for perjury or shading of testimony for the prosecution must be recognized.
Id. at 95 (emphasis omitted). One defense attorney, in testimony to the Grand Jury, described the system as being like
a ‘secret society1 where even though nothing is said, the prosecutors and the informants know that some benefit will flow to the informant for his testimony. The defense attorney explained it was extremely difficult to try to impeach jail house informants when there was nothing in the record relating to benefits they were to receive from their testimony.
Id. at 39.
The Grand Jury concluded that the Los Angeles District Attorney’s office was aware of abuses concerning jailhouse informants well before Leslie White’s 1988 revelations. Evidence supporting this conclusion went back to the late 1970s. Id. at 97. The Grand Jury was particularly critical of the District Attorney’s office’s failure to keep any database of jailhouse informants. Id. at 105-117. It noted that in the high profile cases in which twenty informants would claim to have heard a confession, the informants “were narrowed to a very few, sometimes to only one or two witnesses. No record was kept of those who were rejected and the reasons for that determination. If the informants were again to offer testimony in other cases, there existed no systematic means to review the knowledge obtained by the earlier prosecutor.” Id. at 111.
*1008The Grand Jury report discussed a proposal to create a repository of very basic information about jailhouse informants. The District Attorney’s office rejected this proposal. “The reason consistently offered by officials for deciding against the informant system was that defendants might discover information contained in the index.” Id. at 115. The report found that “[njeither a defendant’s rights to know about information affecting the credibility of an informant, nor a prosecutor’s obligation to disclose such information to a defendant, was ever mentioned during the discussion of the pros and cons of an informant index, according to all sources of evidence presented to the Grand Jury.” Id. at 117.
The Grand Jury reached two overarching conclusions:
A. The Los Angeles County District Attorney’s Office failed to fulfill the ethical responsibilities required of a public prosecutor by its deliberate and informed declination to take the action necessary to curtail the misuse of jail house informant testimony.
B. The Los Angeles County Sheriffs Department failed to establish adequate procedures to control improper placement of inmates with the foreseeable result that false claims of confessions or admissions would be made.
Id. at 6.
The Grand Jury discussed potential relief for defendants, like Gonzales, who had been convicted based on informants’ potentially false testimony during the period covered by the report. “Because the judgments are final, the review must be sought by Petition for Writ of Habeas Corpus. In filing such petitions, certain factual allegations must be made which are legally sufficient to support the relief sought. It is difficult and in some cases likely not possible to allege sufficient facts without discovery.” Id. at 152. The Grand Jury noted that the question of whether defendants could get such post-judgment discovery was currently before the California Supreme Court. Id. The Grand Jury was almost certainly referring to Gonzales’s case which was then pending in that Court. The Grand Jury noted that the District Attorney’s office had been cooperative during the investigation. It expressed the hope that “[i]n the event post-conviction discovery is denied by the [California] Supreme Court ... the District Attorney will be equally cooperative in considering the needs of affected parties for the information necessary to pursue their remedies.” Id. at 153.
C. State Court Proceedings After the Grand Jury Report
The Grand Jury’s hope went unfulfilled. As I describe below, the District Attorney’s office did not remain “equally cooperative” in Gonzales’s case, and the California Supreme Court denied post-conviction discovery.
Gonzales was sentenced to death on May 7, 1981. Shortly thereafter, Gonzales filed a habeas petition in the California Supreme Court. He filed an amended habeas petition on June 10, 1986. Between the time of his death sentence and his habeas filings, Gonzales had been able to obtain almost no additional information concerning Acker. He raised a Brady claim in his state habeas petition, but his discussion of the claim was brief, occupying only six of seventy-seven pages. In this initial habeas petition, Gonzales claimed that the prosecutor had not disclosed the extent of Acker’s cooperation with the police in other cases. Gonzales argued that this violated Brady because such information would have allowed him to argue that Acker’s
*1009testimony violated the Sixth Amendment based on Massiah v. United States, 377 U.S. 201, 84 S.Ct. 1199, 12 L.Ed.2d 246 (1964), and United States v. Henry, 447 U.S. 264, 100 S.Ct. 2183, 65 L.Ed.2d 115 (1980). See People v. Gonzalez, 51 Cal.3d 1179, 275 Cal.Rptr. 729, 800 P.2d 1159, 1192 (1990).
In October 1988, after Gonzales filed his amended state habeas petition, Leslie White triggered the informant scandal and subsequent Grand Jury investigation. The information that came out during the investigation suggests strongly that Acker was involved in the schemes described in the Grand Jiffy report. First, Acker’s informing fit the time frame of the report. The report covered 1979 through 1990. Gonzales’s case, in 1979, was the first case in which Acker gave information concerning an alleged confession. Acker then gave information about five more alleged confessions over the next two years. Second, Acker was placed near Gonzales for two weeks shortly after Gonzales was arrested. Gonzales was arrested on May 29, 1979. He had been shot and taken from the scene in an ambulance, then later taken to the jail. His conversations with Acker took place in early July 1979. This sequence of events is consistent with the Grand Jury’s finding that the Sheriffs Department placed informants near inmates from whom they wanted a confession. Third, the Sheriffs Department would have been particularly interested in a confession from Gonzales because he had shot a Sheriffs Department deputy. Fourth, the Grand Jury found that jailhouse informants were more likely to have allegedly heard confessions in death penalty cases. Finally, Acker claimed to have been helping Gonzales check the accuracy of his attorneys’ work when Gonzales allegedly confessed to him. The Grand Jury found that informants would offer to assist inmates, based on their purported legal experience, in order to obtain information about the inmates’ crimes. Grand Jury Report at 31.
Starting in November 1988, while the Grand Jury investigation was still ongoing, the District Attorney’s office began to address its use of perjured testimony by jailhouse informants. In various memoranda, the office indicated that any problems with the reliability of past jailhouse informants would be revealed to the state courts. On November 17, 1988, Chief Deputy District Attorney Gregory Thompson sent a memorandum to the Deputy District Attorneys, alerting them to the pending review of the office’s use of jailhouse informants. With respect to informants who had been used in the past, Thompson wrote that “the most objective way to do this is on a case by case basis before the court. This will insure an independent review on the merits of each case.” On November 30, 1988, in a letter to various defense attorneys, including Gonzales’s attorney, Thompson again emphasized his intention to make information obtained available through court proceedings:
[W]e planned from the outset to make the information we obtained [through internal investigations] available to defense counsel on individual cases and anticipated that, in appropriate cases, that information would be thoroughly aired in open court.... The best forum for an objective, thorough examination of the issues is in open court.... Further, the court is the only forum where any wrongs can be righted, where any breach in procedure can be assessed, where any prejudice can be weighed, where any injustice can be addressed. It is our aim that every pertinent, relevant issue on each contested case receive the careful and thorough attention of the court.
*1010Thompson also sent letters specifically addressed to attorneys representing clients in whose trials jailhouse informants had testified. The letter to Gonzales’s attorney informed him of the investigation into the use of informants. The letter concluded, “[S]inee the courtroom is the appropriate forum in which to fully explore this issue, our office will join with you in expediting the hearing of any appropriate motion you may wish to bring.” Based in part on this letter, on March 29, 1989, Gonzales’s attorneys filed a 77-point discovery motion with the state court judge who had presided over Gonzales’s trial. The motion sought discovery of information that would reveal whether the prosecution had violated its obligations under Brady to disclose information concerning Acker. Among the information requested was, “Any and all California Department of Corrections Health and Welfare Agency records pertaining to WILLIAM GERALD ACKER.” The primary basis for the motion was the information that had come to light during the Grand Jury investigation of the informant scandal.
The state court trial judge held an initial hearing on Gonzales’s discovery motion in June 1989. Despite the Los Angeles County District Attorney’s repeated prior statements that it wanted to air in open court whatever information it possessed concerning jailhouse informants, it now argued that because there was no proceeding currently before the trial court, that court had no jurisdiction over the case. In response, the judge noted the “very unique factual situation that’s involved in this case.” He said that “if all of the information is true with respect to the allegations that have been made by [Gonzales’s attorneys], I think it puts this case in an entirely different posture.” He stated that “fundamental fair play in this case would indicate that [Gonzales] ought to get that information.” In a second hearing in August, the trial judge granted the discovery motion. He made clear his belief that “the District Attorney’s Office has got to accept much of the responsibility for what occurred, in the sense that ... at no time did your office ever maintain any coordinated effort ... so that the problems which arose could have been prevented.”
The state filed a writ of mandate with the California Supreme Court, seeking to reverse the trial judge’s discovery order. The state argued not only that the judge had no jurisdiction to issue the discovery order, but also that “Gonzalez did not make the showing necessary for a grant of discovery.” The California Supreme Court consolidated the writ of mandate appealing the trial judge’s discovery order with Gonzales’s habeas petition then pending in the Supreme Court.
Gonzales offered to drop his discovery motion if a representative of the District Attorney’s office would swear under penalty of perjury that the state’s files contained no material evidence relating to Acker’s testimony. The District Attorney’s office did not accept the offer. Gonzalez, 275 Cal.Rptr. 729, 800 P.2d at 1219 n. 8 (Broussard, J., dissenting). Gonzales’s habeas attorney wrote in one of her filings to the California Supreme Court, referring to Thompson’s November 30, 1988, letter, “For close to one year, Counsel in this case as well as other cases, all of which are in differing procedural postures, have attempted to ascertain the name of the mysterious but ‘appropriate motion’ that the District Attorney would join in. These efforts have met with overwhelming silence.”
The California Supreme Court granted the state’s writ of mandate, reversing the trial judge’s discovery order. It concluded that “the trial court lacked jurisdiction to *1011order ‘free-floating’ post-judgment discovery when no criminal proceeding was then pending before it.” Gonzalez, 275 Cal. Rptr. 729, 800 P.2d at 1203. The Supreme Court also refused to grant discovery in the pending state habeas petition Gonzales had filed directly in the Supreme Court. Id., 275 CaLRptr. 729, 800 P.2d at 1205. The Court wrote that a habeas corpus proceeding does not “trigger a right to unlimited discovery.” Id. A habeas petition “must set forth specific facts which, if true, would require issuance of the writ. Any petition that does not meet these standards must be summarily denied, and it creates no cause or proceeding which would confer discovery jurisdiction.” Id. Even treating the discovery request and grand jury report as an amendment to the habeas petition, the Court concluded that Gonzales had failed to set forth facts requiring the issuance of the writ. It wrote:
At most, there are new indications that during the period 1979-1988, various residents of the Los Angeles County jail developed successful schemes for fabricating jailhouse confessions by other inmates, and that these practices were ignored or even encouraged by the authorities. However, nothing in the materials before us identifies Acker as a participant in the schemes alleged. Nor is there any specific indication that the prosecution’s files would yield information that substantially undermines Acker’s testimony.... [T]here is no post-conviction right to “fish” through official files for belated grounds of attack on the judgment, or to confirm mere speculation or hope that a basis for collateral relief may exist.
Id. (footnotes omitted).
The California Supreme Court thus concluded that despite the revelations of the District Attorney’s massive use of perjured testimony of jailhouse informants; despite the District Attorney’s failure to establish any database system concerning jailhouse informants for fear that such a system would undermine the effectiveness of their testimony at trial; despite the similarities between Acker’s story and the schemes detailed in the Grand Jury report; and despite the state’s assurance in its letters to defense attorneys that it wanted any problems concerning jailhouse informants to be aired in court, Gonzales’s attempt to discover potential Brady material concerning the jailhouse informant’s testimony in his case was a fishing expedition. The Court ended its opinion: “We expect and assume that if the People’s lawyers have [Brady] information in this or any other case, they will disclose it promptly and fully.” Id., 275 Cal.Rptr. 729, 800 P.2d at 1206.
D. Psychiatric Reports
The Court’s stated expectation and assumption were wrong. The state disclosed nothing to Gonzales voluntarily. Gonzales finally obtained a discovery order in his habeas proceeding in federal court. Under compulsion of the federal court order, the state finally revealed previously undisclosed information about Acker. The most damning information was contained in six psychiatric reports prepared while Acker was incarcerated in California prisons. If the California Supreme Court had allowed the discovery that Gonzales had sought— and indeed had been granted by the trial court — these reports would have been in the state court record on state habeas.
These psychiatric reports could have been used to impeach Acker in three ways. First, they demonstrate that Acker was willing to lie and manipulate prison medical staff in order to get what he wanted. Psychiatric evaluations reveal that Acker admitted attempting or faking suicide on three occasions in order to obtain prison *1012transfers. A 1972 Psychiatric Evaluation reported that Acker admitted that he had faked suicide at the Norwalk Receiving Center in order to be placed in the hospital, from which it would be easier to escape. A 1973 Psychiatric Evaluation reported an attempted hanging, this time in the California Men’s Colony (“CMC”) East Facility. The Evaluation noted that Acker “stated that that was only a gesture designed to prevent his egress from CMC East.” Finally, a 1974 report describes another suicide attempt in April or May of that year at K-Wing of the Duel Vocational Institution (“D.V.I.”). Acker stated “that he made a suicidal gesture in order to get out of K-Wing at D.V.I.” He claimed that “his suicidal gesture was the only way that he could get out of D.V.I.’s K-Wing. Claims that if he ever goes back [to] that madness again at D.V.I. he will attempt the same thing again; namely, a suicidal gesture.”
Many of the psychiatric evaluations specifically noted Acker’s manipulative behavior. One psychiatrist reported that Acker asked, “and it was my distinct impression that this was made in all seriousness, as to ‘how much money would it take to give me a good report.’ ” Another psychiatrist described Acker as “the type of an individual who was constantly testing me.” In a diagnostic study prepared for Acker’s sentencing for his murder conviction, Dr. Flanagan, who examined Acker, recommended that Acker be “considered capable of any measure of brutality in the service of achieving what he wants to do.” He further described him as “intelligent, manipulative, [and] unscrupulous.” The Social Evaluation included in the diagnostic study noted that Acker had been “offered therapy in the past and has used it as a tool to get out of prison early and has not attempted to use it as a tool to change his own behavior.” This evidence of Acker’s lying and manipulation would have been extremely valuable to Gonzales, as Bencangey had no examples during trial of Acker either lying or manipulating to his own advantage. The evidence shows the extent to which Acker was willing to go to obtain prison transfers, the precise reason he testified he was informing against Gonzales. Jurors could easily have concluded that if Acker was willing to repeatedly fake suicide to obtain a prison transfer, he would be willing to fabricate a confession to obtain such a transfer.
Second, these reports could have been used to impeach Acker on his claim that he had started testifying because he wanted to achieve a new “moral balance.” As I describe above, Acker had testified that one of his primary motivations for giving information to the authorities in Gonzales’s case was to turn his life around and to do the right thing. In 1972, Acker had told Dr. Land that he was now on the right path because he had “undergone a religious experience” and he “believes he has found the true religion.” After this claimed conversion, Acker was released from prison. Acker then committed first-degree murder and numerous robberies. That Acker had previously lied about turning his life around in order to obtain his release from prison, and that in fact he had not turned his life around, would have substantially undermined his claim at Gonzales’s trial that he was testifying because he wanted to achieve a new “moral balance.”
Third, these reports could have been used as evidence of Acker’s mental illness. Acker was diagnosed with “schizophrenia, chronic, undifferentiated type” as early as 1974. In 1975, he was diagnosed with “[s]chizophrenia, residual type.” A 1977 report noted that Acker “has a severe personality disorder, and he has previously been diagnosed as schizophrenic.” It further described him as “mentally unstable.” *1013We have repeatedly recognized the impeachment power of mental illness. See, e.g., Silva v. Brown, 416 F.3d 980 (9th Cir.2005).
During discovery for his federal habeas claim, Gonzales was allowed to take Acker’s deposition. Acker’s response to being confronted with these reports in his deposition suggests how he would have responded on the witness stand during trial. First, Acker repeatedly denied making suicidal gestures to obtain transfers and denied bribing prison psychiatrists. When confronted with the report from Dr. Land that he had admitted to having faked suicide in order to be placed in the hospital, Acker responded, “I don’t even know this dude,” and denied that this had occurred. When confronted with another report about the same incident, Acker described the report as “a totally bogus statement right there.” When confronted with the report in which he had admitted that his suicidal gesture was the only way to get out of D.V.I.’s K-Wing, Acker denied making the statement, saying, “That never happened, man.” When confronted with another report about the K-Wing incident, Acker said, “I don’t think I ever said that. And if I did, I lied.” Finally, in responding to Dr. Malloy’s report, which recounted Acker’s attempt to bribe him, Acker denied the incident, describing the report as “full of shit, man.” Second, Acker attacked the psychiatrists who evaluated him. When asked whether Dr. Flanagan was lying when he stated that Acker would do anything to get what he wanted, Acker responded, “Dr. Flanagan is full of crap.” He further stated, while Gonzales’s federal habeas attorney was preparing a question, “If they’re psychiatrist things, don’t go there, man, because they’re all mostly bogus.”
Acker’s psychiatric reports, combined with his response to those reports, would have severely undermined Acker’s credibility at trial. The reports would have revealed that obtaining a prison transfer was so important to Acker that he was willing to fake suicide, manipulate medical professionals, and bribe a prison psychiatrist. Acker’s responses to these reports on the stand would have revealed Acker as a liar, consistently denying events reported by prison psychiatrists.
Gonzales, however, did not have these reports during trial, for the state had successfully concealed them. Gonzales also did not have these reports during his state habeas proceeding, for the state, assisted by the ruling of the California Supreme Court, had again successfully concealed them.
E. Pmholster
I believe we can consider the six psychiatric reports consistent with the Supreme Court’s decision in Pmholster. I would hold that when a petitioner’s inability to present Brady evidence to the state courts is due to the refusal of the state court to allow appropriate discovery, Pinholster does not bar federal courts from considering that evidence in the first instance.
Pinholster does not answer the question whether federal courts can consider evidence that a petitioner tried to discover in state court, but was prevented from discovering by the state court. Justice Soto-mayor stated in her dissent in Pinholster, “I assume that the majority does not intend to suggest that review is limited to the state-court record when a petitioner’s inability to develop the facts supporting his claim was the fault of the state court itself.” 131 S.Ct. at 1417 n. 5. Justice Soto-mayor cited the oral argument transcript in Bell v. Kelly, a case in which this question was presented, but in which the Supreme Court dismissed certiorari as improvidently granted. Bell v. Kelly, 553 *1014U.S. 1031, 1031, 128 S.Ct. 2108, 171 L.Ed.2d 228 (2008) (granting certiorari on question of whether § 2254(d) applies to a claim predicated on evidence the state court refused to consider and that was properly received for the first time in an evidentiary hearing on federal habeas); Bell v. Kelly, 555 U.S. 55, 55, 129 S.Ct. 393, 172 L.Ed.2d 353 (2008) (dismissing certiorari as improvidently granted). The majority in Pinholster never contradicted Justice Sotomayor’s assumption.
The state-court procedural history in Pinholster bears no resemblance to this case. Pinholster made an ineffective assistance of counsel claim in his state habeas petition. Pinholster, 131 S.Ct. at 1396. He relied primarily on the medical diagnosis of a single psychiatrist. Id. On federal habeas, Pinholster presented evidence from two new medical experts who diagnosed him with organic personality syndrome and partial epilepsy and brain injury. Id. at 1397. Pinholster could have presented the opinions of these experts to the state court. Indeed, the question presented in Pinholster was “[wjhether a federal court may reject a state-court adjudication of a petitioner’s claim as ‘unreasonable’ ... based on a factual predicate for the claim that the petitioner could have presented to the state court but did not.” Pet. for Writ of Cert., Pinholster, 131 S.Ct. 1388. The focus of the oral argument in Pinholster was on evidence that could have been presented in state court. See, e.g., Transcript of Oral Argument at 29, Pinholster, 131 S.Ct. 1388 (Justice Kennedy: “Was it the court relied on different evidence, evidence that was not in the State hearing? And that’s the question, whether or not they can do that, if this evidence could have been presented. And certainly it could have been presented.”); id. at 38 (Justice Alito: “The factual predicate of the claim is the new evidence that’s brought forward in ... the Federal proceeding, and unless there is a good reason why that wasn’t brought forward in the State proceeding, it shouldn’t be considered.”).
Based on the revelations of the Los Angeles District Attorney’s use of perjured testimony, Gonzales knew during his state habeas proceedings that there was a strong probability that there was undisclosed Brady material about Acker. Gonzales also had letters from the state effectively inviting him to file a discovery motion seeking such material. The state trial judge who had presided over Gonzales’s trial granted his post-trial discovery motion. But Gonzales was opposed on all fronts by the state, and rejected on all fronts by the California Supreme Court. Unlike Pinholster, who could have put the relevant evidence before the state court but failed to do so, Gonzales tried every means possible for putting the evidence before the state court, but was prevented from doing so by the combined actions of the prosecutor and the state Supreme Court.
As Judge Clifton’s opinion in this case notes, some of the language in Pinholster could be read to cover this case. However, given that the language in Pinholster goes beyond the question before the Court, and given the dramatic differences between Pinholster and this case, I would not read the language broadly to cover the case before us. “It is a maxim not to be disregarded, that general expressions, in every opinion, are to be taken in connection with the case in which those expressions are used. If they go beyond the case, they may be respected, but ought not to control the judgment in a subsequent suit when the very point is presented for decision.” Cent. Va. Cmty. College v. Katz, 546 U.S. 356, 363, 126 S.Ct. 990, 163 L.Ed.2d 945 (2006) (quoting Cohens v. Vir*1015ginia, 19 U.S. (6 Wheat.) 264, 399-400, 5 L.Ed. 257 (1821)). I therefore read Pin-holster as leaving open the question whether federal courts can consider new evidence in a § 2254(d)(1) case when the petitioner’s failure to introduce the evidence in state court was, as here, the “fault of the state court itself.”
I would hold that we can consider such evidence. A longstanding rationale for excluding new evidence is that federal courts should not reward petitioners whose failure to introduce evidence in state court was due to their lack of diligence. This diligence requirement is codified in 28 U.S.C. § 2254(e)(2), which generally bars the introduction of evidence on federal habeas if the petitioner “has failed to develop the factual basis of a claim in State court proceedings.” The Supreme Court has held that a petitioner has not “failed to develop” the factual basis of a claim when the petitioner was “not at fault” for the failure. See Holland v. Jackson, 542 U.S. 649, 652-53, 124 S.Ct. 2736, 159 L.Ed.2d 683 (2004) (citing Williams v. Taylor, 529 U.S. 420, 431-37, 120 S.Ct. 1479, 146 L.Ed.2d 435 (2000)). Before Pinholster, courts of appeal interpreted the diligence requirement in § 2254(e)(2) as the only limitation on the introduction of new evidence in federal court. See Pinholster, 131 S.Ct. at 1417 (Sotomayor, J., dissenting) (“The majority charts a ... novel course that, so far as I am aware, no court of appeals has adopted.”). So long as the petitioner had satisfied the requirements of § 2254(e)(2), courts of appeal had held that federal courts could consider new evidence even for a claim governed by § 2254(d)(1).
In Pinholster, the Court emphasized a different rationale than lack of diligence by the petitioner for refusing to consider evidence presented for the first time to a federal court. The Court held that for a federal court to conclude that a state court “unreasonably] appli[ed]” federal law based on evidence that was not before it is to bypass the state court’s decision-making process. The Court’s holding was “compelled by the broader context of the statute as a whole, which demonstrates Congress’ intent to channel prisoners’ claims first to the state courts.” Pinholster, 131 S.Ct. at 1398-99 (internal quotation marks and citations omitted). It was based on prior cases holding that “review under § 2254(d)(1) focuses on what a state court knew and did.” Id. at 1399.
The Court’s interpretation of the “unreasonable application” language as barring the introduction of new evidence in federal court makes sense when the petitioner’s failure to introduce the evidence in federal court was unrelated to any defect in the state court process. However, when a petitioner’s failure to introduce evidence before the state court is due to the state court’s unwarranted refusal to allow the petitioner to discover the evidence, consideration of the new evidence by the federal court is appropriate. The state court had the opportunity to obtain and consider the evidence, but chose not to do so.
In this case, the state court was presented with the same well-founded discovery request that, in federal court, revealed the undisclosed Brady material. In ruling on Gonzales’s discovery request, the state Supreme Court knew that the Los Angeles District Attorney’s office had been regularly using the testimony of jailhouse informants who were unreliable and had given perjured testimony. It knew that Acker was a jailhouse informant whose story was very similar to the stories of many of the perjuring informants discussed in the Grand Jury report. It knew that the District Attorney’s office had told Gonzales’s counsel, as well as the counsel for other defendants, that “the courtroom is the ap*1016propriate forum in which to fully explore” issues relating to jailhouse informants, but that the office had then opposed all discovery requests. And it knew that the judge who had presided over Gonzales’s trial, and who knew the case intimately, had been persuaded to grant discovery.
The California Supreme Court nevertheless blocked Gonzales’s attempts to obtain discovery to investigate Acker’s reliability. As Justice Broussard noted in his dissent to the Supreme Court’s decision, “whatever defendant tries, the majority find a technical barrier.” Gonzalez, 275 Cal. Rptr. 729, 800 P.2d at 1219 (Broussard, J., dissenting).
Specifically, the majority conclude that (a) defendant cannot get discovery in connection with his automatic appeal because the appeal is limited to the appellate record; (b) he cannot get discovery in an independent trial court action because discovery must be ancillary to a pending proceeding; (c) he cannot get discovery in connection with his pending habeas corpus petition because the issue of Acker’s perjury is not within the scope of the order to show cause; and, (d) even taking into account new allegations of Acker’s perjury and new evidence which gives plausibility to those allegations, he cannot obtain discovery by filing a new habeas corpus proceeding alleging Acker’s perjury because, without discovery, he cannot allege sufficient facts proving Acker’s perjury to state a prima facie case. In short, the majority pose a perfect “Catch — 22” logical conundrum, under which defendant cannot obtain discovery of the law enforcement records because without the information contained in those records he cannot file an action that would support discovery.
Id. In this circumstance, when the state court was faced with the same well-founded discovery request later presented to the federal court, it does not serve AEDPA’s goals to send the diligent petitioner back to the very state court that blocked his attempt to discover the evidence in the first place.
Reading Pinholster to apply to a case where petitioner’s failure to introduce the relevant evidence in state court is the fault of that court is inconsistent with one of the traditional purposes of federal habeas relief, which is to remedy inadequate state court factfinding. In Townsend v. Sain, the Supreme Court held that a federal district court must grant an evidentiary hearing to a habeas applicant if “the fact-finding procedure employed by the state court was not adequate to afford a full and fair hearing.” 372 U.S. 293, 313, 83 S.Ct. 745, 9 L.Ed.2d 770 (1963), overruled on other grounds by Keeney v. TamayoReyes, 504 U.S. 1, 5-6, 112 S.Ct. 1715, 118 L.Ed.2d 318 (1992). The “adequacy of a state-court procedure under Townsend is largely a function of the circumstances and the interests at stake. In capital proceedings generally, this Court has demanded that factfinding procedures aspire to a heightened standard of reliability.” Ford v. Wainwright, All U.S. 399, 411,106 S.Ct. 2595, 91 L.Ed.2d 335 (1986). To hold that an improper state court denial of discovery necessary to develop a federal constitutional claim prevents a federal court from considering in the first instance evidence discovered during federal habeas unnecessarily binds the federal court to the inadequate factfinding of the state court.
I would therefore hold that Pinholster does not bar our consideration of the evidence discovered by Gonzales. I would not reach the question of whether the Brady claim Gonzales raised in federal court is sufficiently distinct from the claim he raised in state court to constitute a “new claim.” See Pinholster, 131 S.Ct. at 1401 *1017n. 10; id. at 1417-19 (Sotomayor, J., dissenting).
Conclusion
I folly concur in the general application of Rhines to cases in which new evidence is introduced on federal habeas. However, for the reasons I have just given, I believe that in this case we have the authority, consistent with Pinholster, to decide Gonzales’s Brady claim now. I would prefer to exercise that authority.
. Given the Court’s deliberate use of the term "hold” and the centrality of the question to the case, I am perplexed by the concurrence’s attempt to dismiss this language as dicta. There simply is no open question of "what federal courts should do when confronted with ... evidence” not yet presented to the state court. Op. at 999. If "evidence introduced in federal court has no bearing on § 2254(d)(1) review,” Pinholster, 131 S.Ct. at 1400, it has no influence in federal habeas petitions.