FACTS AND PROCEDURAL HISTORY

1. Historical Context Facts1

In the early morning hours of November 14, 1984, workers at a restaurant located inside Los Angeles International Airport found the body of the restaurant’s night supervisor, Josephine E. DeBaun. DeBaun suffered two fatal stab wounds – one to the abdomen and one to her neck. The neck wound slashed DeBaun’s carotid artery and nearly severed her head from her neck. She also had nonfatal defensive wounds to her hands, indicating she had resisted her attacker. DeBaun’s body was found in the restaurant’s “money room,” a two-foot by eight– foot enclosed cage in the restaurant’s back office. About $2,100 was missing from the room. The wound to DeBaun’s carotid artery caused blood to spurt around the room, landing on the countertop, filing cabinet, walls and floors. Bloody palm and shoe prints were found in the money room. A blood-stained cloth was found beneath an open first- aid kit, from which some bandages were missing. As a result, the police believed that DeBaun’s attacker had been injured as DeBaun defended herself. It was generally accepted that only one person carried out the attack because the money room was too small to hold more than two people at a time. Henry Earl Duncan was a cashier at the restaurant and had worked earlier that night. After his shift ended at 11:00 p.m., an airport custodian saw someone who looked like Duncan standing on a mezzanine overlooking the restaurant. Duncan was supposed

1 These facts come mostly from our previous decision in this case, People v. Duncan (Aug. 1, 2012, B230459 [nonpub. opn.]).

2 to work the night after the murder but did not show up for his shift. He returned to work two days later. The police interviewed him, but saw no signs that he was injured. Three months later, Duncan stole $2,070 from the restaurant. When he was arrested one month later, the police found under his bed gym shoes with a sole pattern that was consistent with the bloody shoe prints found on the money room floor. Inside Duncan’s car the police also found a key to the restaurant’s cash box. Expert testimony also identified bloody palm and fingerprints found at the murder scene as his. Duncan’s theory at trial was that, although he took part in the theft, someone else committed the murder. Police Criminologist Greg Matheson testified that most of the blood samples tested from the crime scene were Type O, the same as victim DeBaun. A few other samples contained Types A and B as well, indicating that the killer had been injured during his struggle with DeBaun and had blood type A, B, or AB. Duncan’s defense lawyer never ordered a test of Duncan’s blood, which, as it turned out, was also Type O. Duncan was convicted of robbing and murdering DeBaun, along with a special circumstance allegation that the murder occurred during the commission of a robbery. (Pen. Code, §§ 12022, subd. (b), 190.2, subd. (a)(17).) Based on the special circumstance allegation, Duncan was sentenced to death. The judgment was affirmed in People v. Duncan (1991) 53 Cal.3d 955 (Duncan I). Duncan then sought post-conviction habeas corpus relief in the federal courts, contending that he received ineffective assistance of counsel because his lawyer failed to develop and introduce the blood type evidence, which would have shown that someone other than Duncan had attacked DeBaun. In Duncan v. Ornoski (9th Cir. 2008) 528 F.3d 1222 (Duncan II), the Ninth Circuit affirmed the robbery and murder convictions because, among other things, sufficient evidence showed that Duncan had been present while those crimes occurred. The court granted the habeas petition as to the special circumstance finding because Duncan’s trial lawyer had been ineffective by failing to determine whether Duncan’s blood matched that found at the crime scene. Had the jury learned that the blood of a type different from both Duncan’s and DeBaun’s had been found at the scene, the jury would have likely

3 found a reasonable doubt that Duncan had been the actual killer. Given the absence of evidence that Duncan intended that DeBaun be killed, the Ninth Circuit concluded that the death penalty special circumstance finding had to be vacated. (Id. at pp. 1245-1246.) The matter was remanded for a new trial in 2010 on the special circumstance allegation only. At the retrial, the prosecution sought a sentence of life without parole instead of the death penalty. The jury found true the special circumstance that Duncan murdered DeBaun during the commission of a robbery, and the trial court sentenced him to life without possibility of parole, plus four years. We affirmed the judgment in People v. Duncan ((Aug. 1, 2012, B230459) [nonpub. opn.] (Duncan III).) At the second trial, Duncan’s new defense counsel did not introduce the blood type evidence that led the Ninth Circuit to overturn the earlier conviction. Nor did she introduce evidence of the missing bandages or of newly developed DNA evidence that also pointed to someone other than Duncan as the killer. In 2012, Duncan filed a habeas petition with this court, contending that he had once again been the victim of ineffective assistance of counsel because his lawyer failed to introduce the blood type, DNA, and missing bandages evidence that would have raised a reasonable doubt whether someone else had killed DeBaun. Duncan’s supplemental petition also alleged that his lawyer’s failure to further investigate the new DNA evidence prevented her from discovering the existence of a more refined DNA testing technique that in fact showed the presence of DNA from a male other than Duncan. We issued an order to show cause why the writ should not be granted.

2. Facts Concerning the Omitted Evidence from the 2010 Retrial

2.1. The Blood Type and Missing First Aid Supplies Evidence At the first trial in 1986, the investigating detective testified that bandages were missing from an open first aid kit, suggesting that “an attacker or attackers . . . injured themselves in the attack.” A police criminologist testified that they collected blood samples from the crime scene to find the blood “of the wounded suspect.” When the police talked to Duncan two days after the crime, they saw no signs of any injury. As

4 mentioned above, the blood typing tests conducted on those samples included both the victim’s Type O blood and the presence of what was most likely Type AB blood, which had presumably been shed by the injured assailant. Duncan’s blood type was O, apparently excluding him as having been the attacker. Type AB blood was on a rag (Item 8) found on a counter below the open first aid kit. All this information was available to Duncan’s new lawyer, Nancy Sperber, who chose not to introduce it in the 2010 retrial. 2.2 DNA Evidence Is Developed for the Retrial In preparation for the retrial, Sperber and the prosecutor agreed for the first time to conduct joint DNA testing of several bloody items from the DeBaun murder scene. Both sides were concerned that testing samples were insufficient to conduct two tests, one by the prosecution and one by the defense. Accordingly, both counsel agreed that Orchid Cellmark laboratory would conduct the testing. Orchid used the Polymerase Chain Reaction (polymerase) method to test the new samples.2 With one exception, Orchid’s March 2009 test results showed that the items contained DNA that either belonged exclusively to DeBaun, or were insufficient to produce a result. The exception came from Item 4, a sample taken from a bloody shoe print. Tests on that sample showed DNA belonging to both DeBaun and an “unknown male,” meaning someone other than Duncan. A March 30, 2010 supplemental report from Orchid added that the results on Item 4 excluded several specific individuals, including Duncan and a man named Darryl Jackson, whom Duncan at one point identified as his accomplice. Sperber sent Orchid’s results to Blaine Kern, the director of Human Identification Technologies laboratory (HIT), to assess their validity. Kern’s October 2010 written report concluded that Orchid’s results were valid. Kern believed that the test results on Items 4 and 1 (blood found near a phone) eliminated Duncan as the source of the male DNA. Kern also concluded that the test results on four other items – Nos. 5, 7, 8, and 10 – were inconsistent with the 1985 blood type analysis because those earlier tests showed

2 We discuss the shortcomings of polymerase testing in section 2 of our Discussion, and explain why the use of this testing method was ineffective assistance of counsel in section 3 of our Discussion. 5 blood of a type belonging to neither DeBaun or Duncan, which eliminated Duncan as a source, while the recent DNA testing showed only DeBaun’s DNA. Sperber chose not to introduce this new evidence as well.

3. Sperber’s Rationale for Omitting the DNA, Blood Type, and Missing Bandages Evidence

As noted earlier, Orchid’s DNA tests showed that the blood sample from Item 4, the crime scene shoeprint, included the DNA of two persons: the “major profile” was from victim DeBaun, while the “minor alleles” came from an unknown male, meaning someone other than Duncan. The other items tested included DNA from either DeBaun alone or were insufficient to produce results. The report Sperber received from Blaine Kern, the HIT lab director, agreed with these results, and concluded the Orchid tests were valid. Kern’s report stated that DNA retrieved from blood near a phone eliminated both DeBaun and Duncan as donors, while the DNA from Item 4 came from both DeBaun as the major donor, and someone other than Duncan as the minor donor. As a result, Kern wrote, “Duncan is eliminated as possible DNA donor.” (Original boldface.) Sperber explains her decision not to introduce this evidence in her declaration in opposition to Duncan’s habeas petition. We set forth her two paragraph explanation verbatim: “7. I discussed the conclusions with the specialist at H.I.T. telephonically. I was told that they agreed with Orchid-Cellmark’s findings, and that their procedures were accurate. They also told me that the presence of the small amount of unidentified alleles was insufficient to positively state that another human being was involved in the incident; insufficient to indicate from where the alleles originated; that they could have come from anywhere and/or anyone at any time prior to collection; and, they felt that they could not provide any helpful truthful testimony in the matter.

6 “8. I also spoke with Greg Matheson, who was employed by the Los Angeles Police Department Crime Laboratory.[3] I had other cases wherein Mr. Matheson was called as a prosecution witness and found him to be credible. When I discussed this matter with him, Mr. Matheson indicated that, since the original prosecution, he had become trained in DNA analysis. He further stated that he reviewed his work, and the work of Orchid-Cellmark. I also informed him what H.I.T. had concluded. Mr. Matheson agreed with me that the presence of the unidentified alleles was actually meaningless based on the circumstances of where and how they were found, the unsecure nature of the location and the miniscule amount. He indicated that his testimony would not be able to support the proposition that another male was responsible for the crime.” (Original boldface.) Sperber chose not to introduce evidence that bandages were missing from the money room’s first aid kit “because there was no evidence that there were any bandages there prior to the incident, how many had been there, and/or how many were missing. To ask someone who appears on the scene after the fact, with no knowledge of the condition of the scene prior to the incident, would have invited a properly sustained objection based on lack of foundation. There was no evidence that anyone took bandages from the first aid kit at the time of the incident since there was no evidence of anyone else’s injuries at the scene. All of the blood belonged to the victim. I made a tactical decision to not introduce a ‘smoke screen’ which could undermine the credibility of the defense case.” Sperber’s assertions concerning the usefulness of the DNA evidence also motivated her decision to not introduce the blood type evidence from the first trial. During an in camera hearing at the second trial, Sperber explained why she would hold back both types of evidence: “At the time [of the first trial] there was a guy named Greg Mathison [sic] who did the A, B, O [blood] typing. Apparently it looks like there was somebody else’s blood involved in the crime scene. Subsequent DNA testing by the People as well as defense

3 As noted above, Matheson testified at the first trial. 7 review of that testing indicates that there was no one else’s blood; that there were alleles that could have been from skin cells, saliva, sweat, spit in the totally contaminated crime scene. What we have, based on photographs in 1984 nobody was wearing protective clothing, the People’s DNA tested numerous individuals. And it all came back to the alleles. “It’s for that reason, and Mr. Mathison [sic] was under subpoena to the defense after discussing this with everyone on the defense team meaning my investigator and Mr. Frisco [defense co-counsel] and myself, we have made a tactical decision not to present Mr. Mathison [sic] with the A, B, O typing because it can all be refuted by DNA evidence as well as the defense DNA evidence, meaning if I were to put Mathison [sic] up to say that there appears to be someone else’s blood, that’s not correct. And even Mathison [sic] agrees with me at this point in time. “So as a tactical decision on our part, based on the state of the evidence which is the unknown donor of the foot prints that were there, and that was why we decided not to call Mr. Mathison [sic].” Sperber told the court the following day: “The entire tenor of this case, Judge, change[d] once I received the report from my DNA lab, not [the prosecution’s] lab. [¶] Up until that moment, my witness list had been changing. I had other people, and it was gonna be a much more forensic-orientated case. Because up until that moment, I still believed – and talking to Greg Mathieson [sic] from S.I.D., and I indicated to the court I had done that, that this was going to be a DNA blood case. [¶] Once I got the report from my DNA lab and then called Mr. Mathieson [sic], who was under subpoena to me, that basically blew my whole case. “[¶] “I believe, based on what my lab said . . . and what Mr. Mathieson [sic] said – was that his original A/B/O testing, which indicated antigens from someone other than the victim or my client, were present at this crime scene. “My lab confirmed that these antigens were alleles, not necessarily blood, probably not blood, and were not indicative of someone else being present at the crime

8 scene in view – and they had crime scene photographs indicating that everybody was walking around with no glove, nothing. “They confirmed that you would have to test every single person who went in or out of that place, and the likelihood of that being anybody was slim to none. It then became a very different case.” Sperber stated that she discussed this with Duncan, who agreed with her strategy. Sperber’s account was supported by the declaration of her co-counsel, Charles E. Frisco, Jr. Respondent contends that Sperber was not ineffective for failing to introduce the DNA and blood type evidence because Duncan reportedly told her investigator that he tossed the murder weapon into a vacant lot and that, “if’ someone else had been involved in the crime, that person had not been in the money room. According to the investigator, Duncan told her he was “riding the beef” himself.

4. HIT Lab Director Kern And LAPD Criminologist Matheson Rebut Some of Sperber’s Contentions in Declarations Filed in Connection with Present Habeas Petition

In regard to Sperber’s assertion that LAPD criminologist Matheson told her he had “become trained in DNA analysis” and confirmed for her that the 2009 DNA tests were essentially meaningless and did not show the presence of someone else at the crime scene, Matheson stated: (1) he had “very little recall of having had a conversation in 2010 with an attorney or anyone else about Mr. Duncan’s case” or of what he might have said; and (2) he is not “a technical DNA expert and [has] never performed DNA analysis on casework. I have listened to DNA experts and understand the general concepts of forensic DNA, but I would not hold myself out to be an expert and would not have held myself out to be an expert in any conversation I might have had about this case in 2010.” HIT Lab Director Kern’s declaration stated that his October 2010 report to Sperber included findings and conclusions that were helpful to the defense: the discovery of another male’s DNA in Item 4 and the absence of Duncan’s DNA in the other samples

9 that were tested. According to Kern, Sperber did not ask him to testify at trial and did not consult with him any further about his findings. Based on his records, Kern said he had no contact with Sperber after sending his report.4

5. New DNA Tests Explain the Discrepancies and Point to the Presence of Someone Other Than Duncan in the Money Room

As part of the current habeas proceeding, this court authorized Marc Scott Taylor, the president and lab director of Technical Associates, Inc., to conduct DNA testing on the previously tested crime scene samples. Taylor confirmed that the DNA testing performed in advance of the 2010 retrial employed the polymerase method. When a DNA sample contains genetic material from two people, and less than 10 percent of the material comes from one of them, polymerase testing is far less likely to detect the minor donor’s DNA, leading to a result that shows the DNA of only the major donor. According to Taylor’s declaration, another method, Y-short tandem repeats (Y- STR) testing, was well known in 2009 and 2010 as a form of DNA testing that could determine the presence of male DNA from a minor donor that is mixed with the female DNA of a major donor, thus matching the facts of this case. Taylor conducted Y-STR testing on samples obtained from the blood-stained rag found underneath the first aid kit (Item 8), blood from a shoe print (Item 5), and a cloth square used to remove a blood stain elsewhere on the floor (Item 10). Six cuttings from the Item 8 rag were suitable for testing. Duncan’s DNA was found on portions of those cuttings, some of which were blood stained and some of which were not. Because the stains were most likely human blood, Taylor concluded that

4 While Kern’s and Matheson’s declarations raised factual disputes that might otherwise call for a remand to the trial court to conduct an evidentiary hearing (see In re Vargas (2000) 83 Cal.App.4th 1125, 1131), our conclusion that Duncan received ineffective assistance of counsel is based solely on undisputed evidence. As a result, these evidentiary conflicts play no role in our analysis. 10 Duncan’s DNA came from a source other than blood, such as saliva. Five of the six items that were tested contained the DNA of a man other than Duncan. The test performed on the shoe print blood swab (Item 5) detected low levels of male DNA at three of the 16 genetic loci employed in Y-STR testing. The tests showed two alleles consistent with Duncan, while the third was different. If one man was the donor, then Duncan would be excluded. According to Taylor, the nature of Y-STR testing meant that if two men deposited their DNA in that sample, then, at a minimum, someone in addition to Duncan deposited DNA in the shoe print. Tests performed on the other blood stain (Item 10) were less conclusive, but did show the presence of the DNA of two men. Although it was possible Duncan was the source of some of the male DNA, “it cannot be said that there is a strong likelihood that he was the source.” In short, Taylor concluded, the polymerase testing authorized by Sperber in 2009 was not inconsistent with the blood typing evidence from 1985. Instead, the Y-STR testing he performed merely highlighted the shortcomings of the older polymerase method when testing samples that are known to have DNA from a female major donor and a male minor donor. As a result, the blood typing evidence from 1985 was valid and was further confirmed by the Y-STR testing. Each showed the presence of a third person.

DISCUSSION

1. Principles Applicable to Ineffective Assistance of Counsel Claims

Under the Sixth Amendment to the United States Constitution and article I, section 15 of the California Constitution, a defendant has the right to the effective assistance of counsel at trial, which means the reasonably competent assistance of an attorney acting as his diligent and conscientious advocate. (In re Lucas (2004) 33 Cal.4th 682, 721 (Lucas); see Strickland v. Washington (1984) 466 U.S. 668, 688, 694 (Strickland).) A defendant claiming he received ineffective assistance of counsel must prove by a preponderance of the evidence that: (1) his lawyer’s performance was deficient because it fell below an

11 objective standard of reasonableness in all the circumstances; and (2) that absent those errors, a different outcome was reasonably probable, which means a probability sufficient to undermine confidence in the outcome. (Lucas, at p. 721.) The standards for showing ineffective assistance of counsel are the same at the penalty phase as they are in the guilt phase. (People v. Cunningham (2001) 25 Cal.4th 926, 1030.) The United States Supreme Court has declined to articulate specific guidelines for what constitutes objectively reasonable performance, emphasizing instead that the standard is simply reasonableness under prevailing professional norms. (Lucas, supra, 33 Cal.4th at p. 721, citing Wiggins v. Smith (2003) 539 U.S. 510, 521.) When reviewing counsel’s performance, we presume that counsel’s conduct fell within the “wide range” of reasonably competent assistance. (Lucas, at p. 722.) We must make “every effort . . . to eliminate the distorting effects of hindsight, to reconstruct the circumstances of counsel’s challenged conduct, and to evaluate the conduct from counsel’s perspective at the time.” (Strickland, supra, 466 U.S. at p. 689.) Even though we defer to counsel’s strategic choices, such scrutiny is limited and in certain cases may be unjustified. (In re Jones (1996) 13 Cal.4th 552, 561 (Jones).) Deference is not abdication and may never be used to insulate counsel’s performance from meaningful scrutiny and thereby automatically validate challenged acts or omissions. (Ibid.) What appears to be a tactical decision may actually be unreasonable if based upon inadequate investigation, and counsel must diligently pursue leads that indicate the existence of favorable evidence. (Lucas, supra, 33 Cal.4th at p. 726.) In other words, before counsel acts or chooses not to act, she “must make a rational and informed decision on strategy and tactics founded upon adequate investigation and preparation.” (Id. at pp. 721-722, internal citations and quotation marks omitted.) As the United States Supreme Court observed in Strickland: “Strategic choices made after thorough investigation of law and facts relevant to plausible options are virtually unchallengeable; and strategic choices made after less than complete investigation are reasonable precisely to the extent that reasonable professional

12 judgments support the limitations on investigation. In other words, counsel has a duty to make reasonable investigation or to make a reasonable decision that makes particular investigations unnecessary. In any ineffectiveness case, a particular decision not to investigate must be directly assessed for reasonableness, applying a heavy measure of deference to counsel’s judgments.” (Strickland, supra, 466 U.S. at pp. 690-691.) Finally, because representing an accused murderer is a mammoth responsibility, we also consider the seriousness of the charges when assessing counsel’s performance. (Jones, supra, 13 Cal.4th at p. 566.)

2. A Brief Primer on Y-STR DNA Testing

Y-STR DNA testing is a refinement of the polymerase testing that Orchid had performed in 2009. Polymerase testing looks for genetic markers on 13 loci, or points, along a DNA sample. If a profile from such testing matches that of a sample taken by a particular individual, it is virtually certain that the individual was the source of the sample. (State v. Calleia (2010) 414 N.J.Super. 125, 145-146 (Calleia), reversed on other grounds in State v. Calleia (2011) 206 N.J. 274.) Y-STR testing is the same as Polymerase testing except that it permits analysis of only the male DNA in a mixed source sample that also contains the DNA of a female contributor. (People v. Stevey (2012) 209 Cal.App.4th 1400, 1413 (Stevey), citing Calleia, at pp. 143-145.) Because the male Y chromosome travels throughout a family’s lineage, all men within that lineage will be possible donors of the DNA sample tested under the Y-STR method. (Stevey, supra, 209 Cal.App.4th at pp. 1413-1414.) As a result, such testing cannot identify a particular individual. However, it is extremely useful in excluding someone because an individual cannot be the source of the DNA if the profiles do not match. (Ibid.) That is the result of the new testing performed by Technical Associates, Inc., the lab used in connection with the current habeas proceeding. The DNA of a male other than Duncan appears in the blood samples.

13 3. Duncan Received Ineffective Assistance of Counsel5

Both parties have focused on Sperber’s actions after the Orchid DNA test results came back in 2009. Sperber states in her declaration that, based on what she was told by an unnamed “specialist” at the lab she hired, and by LAPD criminologist Matheson, the failure of the 2009 test results to find the DNA of someone other than Duncan or victim DeBaun made both those results and the 1985 blood type evidence virtually meaningless. Absent evidence that someone else’s blood had been spilled, Sperber also believed that the missing bandages evidence was inadmissible as both speculative and lacking foundation. Duncan contends that Sperber failed to investigate further, especially in light of both the 1985 blood type evidence that showed the presence of blood from a third person and the DNA results for Item 4, which included the DNA of a male other than Duncan. Setting aside the conflicting evidence as to whether Matheson or someone at HIT labs told Sperber the 2009 DNA evidence was useless, Duncan contends that Sperber’s credulous reliance on the comments of a police criminologist and her failure to make further inquiries that should have led her to conduct Y-STR testing fell below the required performance standard. As set forth below, we agree with Duncan’s contentions concerning Sperber’s lack of follow-up investigations. However, we believe the record shows that Sperber also overlooked a more significant issue: whether Sperber ever investigated DNA testing methods before agreeing to joint polymerase testing with the prosecution in the first instance. We begin by returning to the 1985 blood typing evidence, which showed the presence of both Type O blood belonging to DeBaun and Type AB blood, the latter of

5 Because we conclude that Sperber’s omissions concerning the blood type, missing bandages, and 2009 DNA evidence constituted prejudicial ineffective assistance of counsel, we do not reach Duncan’s other contentions concerning Sperber’s failure to call her own crime scene reconstruction expert and her supposedly inadequate closing argument. 14 which was presumed to have come from DeBaun’s attacker as a result of her defensive struggle. Because Duncan was also Type O, and because the police saw no wounds on Duncan when they interviewed him two days after the crime, the presence of Type AB blood strongly suggested that he had not been the actual attacker, although he may have been an aider and abettor. Therefore, the premise for undertaking DNA testing in 2009 was to determine the presence of male DNA in blood samples that contained a mixture of DNA from a female victim and a presumably male attacker. According to DNA expert Taylor, Y-STR DNA testing was uniquely suited for such a situation, and was well known among experts in 2009 and 2010. His declaration to that effect is uncontradicted. Taylor’s contention is supported by other sources. As early as 2004, scientific articles recognized the existence and validity of Y-STR testing. (Calleia, supra, 414 N.J. Super. at pp. 148-149, citing: Sudir K. Sinha et al., Utility of the Y-STR Typing Systems [Y-PLEX 5] in Forensic Casework and 11 Y-STR Haplotype Database for Three Major Population Groups in the United States, 49(4) J. Forensic Sci. 1 (July 2004); Mark A. Jobling & Peter Gill, Encoded Evidence: DNA in Forensic Analysis, 5 Nature Reviews- Genetics 739 (Oct. 2004); Benjamin E. Krenke et al., Validation of a Male-Specific, 12- Locus Fluorescent Short Tandem Repeat (STR) Multiplex, 148 Forensic Sci. Int’l. 1, 2 (2005).) Although Stevey, supra, 209 Cal.App.4th 1400, a 2012 case, is the first reported California decision to explicitly recognize the reliability and admissibility of Y-STR testing, courts across the country began recognizing as early as 2000 that Y-STR testing (and its female counterpart, mtDNA testing) had gained sufficient scientific consensus regarding its reliability to render Y-STR test results admissible. (Id., at p. 1412, and cases cited therein.) Given the facts known to Sperber concerning the possible mixture of male and female DNA in the blood samples to be tested, a reasonably competent lawyer would have consulted with experts or investigated the matter ahead of time to determine that Y- STR testing was the appropriate technique for this case. Sperber’s declaration does not

15 address this issue at all. She does not state what investigation, if any, she did beforehand, or otherwise explain the circumstances surrounding her agreement to submit the samples to standard polymerase testing, which was ill-suited to the task at hand. To determine whether Sperber’s inaction constituted ineffective assistance of counsel, we need to look no further than the Ninth Circuit’s decision in Duncan II, supra, 528 F.3d 1222, which held that Duncan received ineffective assistance of counsel at his first trial because his lawyer failed to investigate the need to conduct blood typing evidence despite police lab tests that showed a mixture of the victim’s Type O blood with the blood of someone else who was most likely Type AB. The crime-scene blood type evidence from the first trial was the only forensic evidence that had not been tied to Duncan, and offered the lawyer “his only opportunity to cast doubt on the State’s theory that Duncan was the actual killer.” (Duncan II, supra, 528 F.3d at p. 1236.) Duncan’s first lawyer never consulted a serology expert and decided not to conduct a blood type test on Duncan because Duncan supposedly told the lawyer he had been in the money room, and the lawyer did not want to risk further tying Duncan to the crime scene. (Id. at p. 1232.) Post-conviction serology tests showed that Duncan had the same Type O blood as the victim, meaning that the Type AB blood mixed in with DeBaun’s could not have come from Duncan. Noting the general rule that counsel must undertake a reasonable investigation before making tactical decisions, the Duncan II court said that “[i]t is especially important for counsel to seek the advice of an expert when he has no knowledge or experience about the field. [Citation].” (Duncan II, supra, 528 F.3d. at pp. 1234-1235.) The lawyer “provided no explanation for why he failed to consult a serology expert or investigate the potentially exculpatory blood evidence. He certainly did not advance a strategic or tactical reason for failing to do so.” (Duncan II, supra, 528 F.3d at p. 1236.) Had he done so, “he would have been prepared to make the necessary decisions regarding the testing of Duncan’s blood, what further investigation should be done, and how to approach the cross-examination of the State’s serology expert. Without consulting an expert, [the first lawyer] had no basis upon which to devise his defense

16 strategy.” (Ibid.) In short, the lawyer’s “failure to consult an expert in serology that would have enabled him to make informed, strategic decisions with regard to the blood evidence” was constitutionally deficient. (Id. at p. 1234.) In Duncan II, supra, 528 F.3d 1222, the Ninth Circuit essentially provided a primer for Sperber to follow at the second trial, but she did not follow through. As the Ninth Circuit noted about Duncan’s lawyer in the first trial, Sperber does not state that she has any expertise in DNA analysis, does not claim that she consulted with any DNA experts before agreeing to testing by the polymerase method, and does not explain why she agreed to do so. Had she performed a proper investigation, she would have been directed to the Y-STR testing method, which would have detected the other male DNA, a result that would have been consistent with the 1985 blood type test results. With substantial evidence on hand that someone else had in fact been the killer, the missing bandages evidence that had been introduced at the first trial would have assumed its proper place in the chain of events supporting Duncan’s theory that someone else killed DeBaun. (Duncan II, supra, 528 F.3d at pp. 1234-1236; see Leonard v. Michigan (W.D.Mich. 2003) 256 F.Supp.2d 723, 729-731 [habeas petition for ineffective assistance of counsel granted where defense counsel failed to consult a DNA expert before unsuccessful evidence suppression hearing].) We therefore conclude that Duncan received ineffective assistance of counsel for Sperber’s inaction regarding the initial DNA tests. We also agree with Duncan that Sperber was ineffective by failing to follow up the initial DNA results with additional investigation and further testing under the Y-STR method. We begin with Orchid’s 2009 test results and HIT’s report to Sperber confirming Orchid’s accuracy. Although most of the samples tested found no DNA other than DeBaun’s, the test results from Item 4 included the DNA of a male other than Duncan, thereby excluding him. Sperber’s declaration states that an unidentified specialist at HIT confirmed that Orchid’s test results were accurate, but also told her the results were insufficient to show that another person’s DNA had been deposited at the crime scene. HIT’s written report was clear however: it stated that Item 4 included the

17 DNA of another male, excluding Duncan as the donor. This created an ambiguity that demanded clarification, but Sperber does not contend that she followed up on this issue. Instead, it appears that she let the matter drop. Her decision to do so appears to have been motivated in part by statements attributed to LAPD criminologist Matheson that the unidentified DNA sample was effectively meaningless given the small amount detected and the supposedly questionable manner of the sample’s collection. Setting aside Matheson’s declaration that he could not recall a conversation with Sperber at that time and was not, and would not have held himself out as, a DNA expert, we find questionable Sperber’s reliance on Matheson’s supposed statements. Such unquestioning reliance is generally considered to be ineffective assistance of counsel. (Elmore v. Ozmint (4th Cir. 2011) 661 F.3d 783, 853-854 [counsel ineffective for failing to conduct independent analysis of state’s forensic evidence and by failing to “mistrust the state’s case” because skepticism of authority “is an absolute necessity for a lawyer charged with representing a client charged with capital murder”]; Anderson v. Johnson (5th Cir. 2003) 338 F.3d 382, 392 [counsel ineffective by relying exclusively on state’s investigative work]; Hash v. Johnson (W.D. Va. 2012) 845 F.Supp.2d 711, 741 [it is unreasonable to rely on findings of a police investigation and not undertake an independent investigation].) Respondent contends that, having performed one test, Sperber was not ineffective for failing to seek out more testing. Respondent supports this proposition with decisions concerning defense counsel’s failure to seek further testing in support of a psychiatric defense theory where the defendant has already been examined by several experts who conclude that no mental condition exists. (People v. Payton (1992) 3 Cal.4th 1050, 1078; In re Fields (1990) 51 Cal.3d 1063, 1074; People v. Williams (1988) 44 Cal.3d 883, 945; People v. Stanworth (1974) 11 Cal.3d 588, 613, overruled on another ground by People v. Martinez (1999) 20 Cal.4th 225, 237.) These decisions have no application here, where the issue is investigation into the need for forensic testing. (See In re Sixto (1989) 48 Cal.3d 1247, 1257, citing Williams, supra, but concluding that trial counsel was

18 ineffective for failing to investigate defendant’s blood alcohol and PCP levels when both were relevant to his mental state].) Finally, to the extent respondent suggests that Duncan’s ambiguous assertions concerning his presence in the money room and other statements that might be construed as admissions of guilt limited the need for further inquiry, that proposition was rejected in Duncan II, supra, 528 F.3d at p. 1238 [“A defendant’s admission of guilt to his lawyer does not absolve the lawyer of his duty to investigate the crime.”].) We therefore also conclude that Sperber rendered ineffective assistance of counsel by failing to further investigate the results of the Orchid lab’s PCR testing.6

4. Sperber’s Omissions Were Prejudicial

In order to establish that he was prejudiced by Sperber’s trial performance, Duncan must show a reasonable probability of a more favorable outcome, meaning a probability sufficient to undermine confidence in the outcome. (Duncan II, supra, 528 F.3d at p. 1239; see Lucas, supra, 33 Cal.4th at p. 721.) This does not require showing that Sperber’s deficient conduct more likely than not changed the outcome. Instead, we examine the evidence that could have been presented to the jury if counsel had performed competently and compare that to the evidence the jury actually heard. (Duncann II, at pp. 1239-1240.) In Duncan II, the Ninth Circuit held that the missing blood type evidence would have raised a reasonable doubt that Duncan had been the actual killer, leaving proof that he at least intended DeBaun’s death as the only avenue for a special circumstances murder finding. (Duncan II, supra, 528 F.3d at p. 1244.) Because there was no evidence regarding Duncan’s intent, and because multiple scenarios concerning his actions and

6 Respondent contends that Sperber was not ineffective because she skillfully cross- examined prosecution witnesses and because she presented expert forensic evidence that a gloved hand left a bloody fingerprint in the money room, which, because Duncan’s palm prints were also found, indicated the presence of an accomplice. We disagree. While this evidence was undoubtedly helpful, it possesses nowhere near the strength of the blood type and DNA evidence. 19 intent were possible, the court held there existed a reasonable possibility that at least one juror would find Duncan acted without the intent to kill. (Id. at p. 1245.) In Duncan III, supra, we also noted that Duncan’s limited evidence concerning the existence of someone else as the actual killer “would have defeated the special circumstance allegation” if the jury had believed it. (Duncan III, supra, slip opn. at p. 10.) In short, as in the first trial, in order to find true the special circumstance allegation that led to a sentence of life without parole, the jury had to find either that Duncan had actually killed DeBaun or intended that an accomplice kill her. Respondent contends there is evidence of both in the trial record. Nearly all the evidence cited by respondent relates to the facts of the crime itself: that Duncan knew the layout, was seen waiting nearby after hours, that the crime was particularly brutal, that shoe prints consistent with Duncan’s shoes were found at the scene, as were his bloody palm and fingerprints, and that a prosecution crime scene expert opined that the only shoe prints found were those consistent with Duncan’s shoes, and how no other palm or fingerprints were found at the scene. Respondent also contends that Duncan’s intent to kill can be shown from the absence of evidence that he did anything to stop the attack or render aid to DeBaun. These arguments, although relevant for jury argument, are not dispositive in the current setting. As Duncan points out, respondent has done no more than show the existence of substantial evidence to support the special circumstances finding based on the evidence presented at the retrial. However, the retrial did not include the DNA, blood type, and missing bandages evidence, and our job is to evaluate that evidence against the scientific evidence that Sperber failed to develop and introduce. Respondent does not address how the blood type and DNA evidence might have changed or not changed the outcome had it been introduced. We believe its effect would have been significant. The Duncan II court held that the absent blood type by itself tended to establish that an accomplice, and not Duncan, killed DeBaun. Had the jury heard that evidence, “there is a reasonable probability that one or more of them would

20 have had a reasonable doubt that Duncan was guilty of the special circumstance alleged.” (Duncan II, supra, 528 F.3d at p. 1244, fn. omitted.) That conclusion becomes even stronger when we factor in not just the 2009 DNA tests showing that at least one item contained the DNA of another male, but the Y-STR evidence that confirms another man was present and that his blood was mixed with DeBaun’s. Given the uncontroverted theory that only one person could have killed DeBaun, along with evidence that a man’s blood was mixed with hers and that bandages were missing from a first aid kit, at least one juror could reasonably conclude that the blood came from the attacker. That Duncan’s palm, finger, and shoe prints might have been found in the money room might definitively show that he was a willing accomplice and that he entered that room, but does not conclusively show when. Given the blood type, DNA, and missing bandages evidence, we believe it is reasonably likely that at least one juror would have rejected the interpretive blood print evidence in favor of the far more precise DNA evidence. As a result, our confidence in a special circumstance finding based on the notion that Duncan had been the killer is seriously undermined. That leaves the possibility that even though Duncan did not carry out the attack on DeBaun, he still intended that she die. As the Duncan II court noted, the evidence from the first trial left open the possibility that Duncan and his accomplice were waiting for DeBaun to leave, and were surprised to find her there, with the accomplice attacking her without Duncan intending that he do so. They could have also entered the restaurant expecting to find her there, with the accomplice adopting the intent to kill when DeBaun put up an unexpectedly strong struggle. (Duncan II, supra, 528 F.3d at p. 1245.) Respondent speculates that Duncan stood by and did nothing to stop the attack or help DeBaun, concluding that such inaction showed his intent to kill. Respondent cites no authority for this proposition, however. At bottom, speculation about Duncan’s conduct or intentions is not evidence and, as in Duncan II, leaves us with insufficient confidence in the outcome. We therefore hold that Duncan was prejudiced by Sperber’s ineffective assistance of counsel.

21 5. Respondent’s Remaining Contentions

Respondent contends that Duncan’s petition is untimely because it was filed more than 21 months after his sentence was imposed at the retrial of the special circumstance allegation. We disagree. Duncan filed his petition on August 22, 2012, just 21 days after our decision in Duncan III, dated August 1, 2012. During the pendency of this proceeding, we authorized further DNA testing, specifically the Y-STR testing that confirmed Duncan had received ineffective assistance of counsel. (See In re Robbins (1998) 18 Cal.4th 770, 780 [petitioner may hold off filing claims while conducting an ongoing investigation].) Much of the delay between the filing of the petition and this opinion is attributed to the district attorney’s inability over a two-year period to provide sufficient samples for defense DNA testing. Samples were not provided until we issued an order to show cause for their production. We therefore conclude that the petition was timely. Respondent also contends that Duncan’s petition relies on inadmissible hearsay from the supporting declaration of Mitchell Zimmerman, who was Duncan’s post- conviction counsel after the first trial in 1986. In that declaration, Zimmerman contends that Sperber rebuffed or ignored certain suggestions he made concerning the proper presentation of Duncan’s case. We express no opinion on the admissibility of Zimmerman’s declaration, and we have not relied on it for our analysis.7

7 Duncan contends that respondent’s return is insufficient because it contains numerous general denials instead of stating facts. (People v. Duvall (1995) 9 Cal.4th 464, 476.) We do not reach this issue, and, to the extent the return might include inadmissible material, we overlook it. 22

DISPOSITION

The petition is granted and the judgment is reversed for a retrial as to only the special circumstances murder allegation.

RUBIN, J.

WE CONCUR:

BIGELOW, P. J.

GRIMES, J.

23