Johnson v. Premo

688 August 3, 2017 No. 39 IN THE SUPREME COURT OF THE STATE OF OREGON MARTIN ALLEN JOHNSON, Respondent on Review, v. Jeff PREMO, Superintendent, Oregon State Penitentiary, Petitioner on Review. (CC 06C16178; CA A154129; SC S064132) On review from the Court of Appeals.* Argued and submitted January 10, 2017. Benjamin Gutman, Solicitor General, Salem, argued the cause and filed the briefs for petitioner on review. Also on the briefs was Ellen F. Rosenblum, Attorney General. Daniel J. Casey, Portland, argued the cause and filed the briefs for respondent on review. Also on the briefs was Robert L. Huggins, Jr. Jeffrey Erwin Ellis, Portland, filed the brief on behalf of amicus curiae Oregon Criminal Defense Lawyers Association. Before Balmer, Chief Justice, and Kistler, Walters, Landau, and Nakamoto, Justices, and Brewer, Senior Justice pro tempore.** BREWER, S. J. The decision of the Court of Appeals and the judgment of the circuit court are affirmed. _______________ **  Appeal from Marion County Circuit Court, Don A. Dickey, Judge. 277 Or App 225, 370 P3d 553 (2016). **  Baldwin, J., retired March 31, 2017, and did not participate in the decision of this case. Flynn and Duncan, JJ., did not participate in the consideration or decision of this case. Cite as 361 Or 688 (2017) 689 Case Summary: Petitioner, who was convicted of aggravated murder and sentenced to death, sought post-conviction relief, alleging that he received inad- equate assistance of counsel in violation of Article I, section 11, of the Oregon Constitution. More particularly, petitioner alleged that his trial counsel failed to adequately investigate the cause of the victim’s death. At the criminal trial, the state presented evidence that petitioner had strangled the victim, and peti- tioner’s defense counsel offered evidence instead that petitioner had drowned the victim but had done so in a county other than the county in which he was tried, and therefore argued that petitioner should be acquitted based on the state’s fail- ure to prove the proper venue. At his post-conviction trial, petitioner presented evidence that the opinions of both the state and defense experts that testified at the criminal trial as to the cause of death were incorrect, and the victim actually died of a morphine overdose. The post-conviction court granted petitioner a new trial, and the Court of Appeals affirmed. Held: Petitioner received inadequate assistance of trial counsel. To be entitled to post-conviction relief based on inade- quate assistance of counsel, a petitioner must show that counsel failed to exercise reasonable professional skill and judgment, and that the petitioner suffered prej- udice as a result of counsel’s inadequacy. Tactical decisions made by counsel must be grounded on reasonable investigation. In this case, counsel knew there was a discrepancy between petitioner’s factual account of what had occurred and what the state and defense forensic experts opined had occurred. Counsel also knew that the medical evidence was complicated, and that there was evidence that the victim had ingested numerous drugs before her death. In that circumstance, and in light of the undesirability of the venue defense, counsel should have sought out additional information concerning the drugs in the victim’s system at the time of her death, in order to try to develop a theory that petitioner’s killing of the victim was unintentional, or alternatively, in the penalty phase that, even though inten- tional, it was not the type of crime for which the death penalty should be imposed. Affirmed. 690 Johnson v. Premo BREWER, S. J. In a two-phased jury trial, petitioner was convicted of aggravated murder and sentenced to death for killing a fifteen-year-old girl, HF. The state’s theory of the case was that petitioner had killed HF intentionally in further- ance of, or in an effort to conceal, the commission of sexual offenses against her. This court affirmed his convictions and sentence. State v. Johnson, 340 Or 319, 131 P3d 173, cert den, 359 US 1079 (2006). Petitioner then brought this action for post-conviction relief, asserting that he received inad- equate assistance of trial counsel in violation of Article I, section 11, of the Oregon Constitution. The post-conviction court granted relief on one ground and denied relief on other grounds, and the Court of Appeals affirmed. Johnson v. Premo, 277 Or App 225, 370 P3d 553 (2016). Respondent sought review in this court. On review, we conclude that the post-conviction court and the Court of Appeals correctly determined that petitioner is entitled to post-conviction relief. At petitioner’s criminal trial, the state presented evidence that petitioner drugged HF with morphine, raped her, strangled her to death, then threw her body off a bridge. Petitioner did not testify. The sole defense theory presented by his trial counsel, Walker and Peters, was that HF had not died by strangulation as theorized by the state, but, instead, had died of drowning after petitioner threw her off the bridge. As a consequence, counsel argued, petitioner was entitled to an acquittal because the state had not initiated the prosecution in the county in which he had drowned HF.1 As noted, that defense was unsuccessful, and the jury con- victed petitioner and sentenced him to death. In this post-conviction proceeding, petitioner asserted, among other things, that the venue defense that his criminal trial counsel advanced had virtually no chance of persuading a jury to acquit him. More specifically, petitioner 1  Venue generally lies in the county in which the offense occurred. ORS 131.305. At the time of petitioner’s criminal trial, venue was considered an ele- ment of a criminal offense. In State v. Mills, 354 Or 350, 312 P3d 515 (2013), this court concluded that venue was not an element of an offense that needed to be found as fact by a jury beyond a reasonable doubt. Cite as 361 Or 688 (2017) 691 argued that, because, based on the evidence in the record, the jury could have found that the place of HF’s death could not be readily determined, a venue defense was not viable in light of the alternative venue provisions of ORS 131.325.2 Moreover, counsel’s sole reliance on such a weak technical defense made the penalty phase of his trial (during which the jury considered aggravating and mitigating factors and determined whether a sentence of death is appropri- ate) much more challenging. Petitioner asserted, instead, that his counsel should have pursued a morphine-overdose theory of the case, in light of petitioner’s statement to his defense team that he woke up after having sex with HF and discovered that she was dead. Petitioner further asserted that, if counsel had consulted a toxicologist, they would have developed credible evidence that HF died of a drug overdose, thus rebutting the state’s evidence that she died by stran- gulation. With respect to the issue of prejudice, petitioner argued that counsel’s failure to pursue a more viable theory of defense that actually conformed to petitioner’s story had a tendency to affect the outcome of his criminal trial. The post-conviction court agreed with petitioner with respect to that claim and, accordingly, granted relief. The Court of Appeals affirmed. The state’s primary argument on review is that defense counsel, having retained an expert who opined that the cause of the victim’s death was drowning, was not required to seek out additional experts to try to establish a cause of death—morphine-overdose—that had been ruled out by both prosecution and defense experts. As explained below, we do not view the relevant inquiry as how many experts should have been consulted; the evaluation of coun- sel’s adequacy is more nuanced than that. The dispositive 2   ORS 131.325 provides: “If an offense is committed within the state and it cannot readily be determined within which county the commission took place, * * * trial may be held in the county in which the defendant resides, or if the defendant has no fixed residence in this state, in the county in which the defendant is appre- hended or to which the defendant is extradited.” The state initially alleged that the crimes had occurred in Washington County, but amended the indictment to allege venue under ORS 131.325. The state pre- sented evidence at the criminal trial that petitioner resided in, and later was extradited to, the county in which he was tried. 692 Johnson v. Premo issue, rather, is whether adequate trial counsel would have attempted to develop a theory of defense that HF already was dead from a drug overdose when petitioner threw her body off the bridge. I. FACTS A.  Evidence Adduced at Petitioner’s Criminal Trial We recount the pertinent facts adduced at peti- tioner’s criminal trial. On February 23, 1998, HF went to petitioner’s house in Washington County to play computer games. Petitioner previously had provided HF with drugs and alcohol, and had expressed sexual interest in her. The following day, HF’s body was discovered on a beach in Clatsop County near the Astoria Bridge at the mouth of the Columbia River. Petitioner fled the state shortly after being interviewed by police concerning his possible involvement in HF’s death, and he was not apprehended for almost a year. At trial, the state presented evidence that peti- tioner had driven to the Astoria area and that HF’s blood was found on his car. The state’s medical examiner, Dr. Hartshorne, opined that HF had died by strangulation, noting fingerprint-shaped bruises on her neck and petechiae on her face. In addition, the state’s evidence showed that HF had a significant amount of morphine in her system, and semen in her vagina matched petitioner’s DNA. The state introduced evidence that petitioner “habitually preyed on underage girls, taking them to nightclubs, providing them with alcohol and drugs, engaging them in consensual sex- ual relations when possible and, most significantly, sexu- ally abusing them while they were rendered unconscious by drugs that he had provided to them.” Johnson, 340 Or at 321.3 Petitioner’s counsel adduced expert testimony from Dr. Ferris, a forensic pathologist who had extensive experi- ence in examining bodies recovered from water. Noting in particular that water and silt had been found in her lungs, 3   Much of the initial focus of petitioner’s defense team had been on trying to exclude evidence of petitioner’s history of drugging and sexually assaulting other teenagers, but that strategy ultimately proved unsuccessful. See 340 Or at 337- 42 (concluding that such evidence was admissible under OEC 404(3)). Cite as 361 Or 688 (2017) 693 Ferris opined that the victim had not died from strangula- tion but, rather, had drowned. Both Ferris and Hartshorne opined that the amount of morphine in the victim’s system was insufficient to have caused her death. Ferris character- ized the amount of morphine in the victim’s body as rela- tively low, although Hartshorne believed that the amount was significant enough that it could have affected the vic- tim’s ability to fight off a strangulation attack. Both Ferris and Hartshorne acknowledged that the victim showed signs of pneumonia from having aspirated vomit, which they con- cluded had occurred when she was unconscious due to the morphine. In sum, there was unrefuted evidence that petitioner had had sexual intercourse with the victim, that the victim had been rendered unconscious due to morphine ingestion, that there were injuries on the victim’s body, including her neck, and that petitioner had thrown her off a bridge. The disputed facts centered on how—and by exten- sion where—petitioner had killed the victim, not on whether he had killed her. In light of Ferris’s testimony, petitioner’s trial counsel argued to the jury that petitioner should be acquitted of aggravated murder because the crime had not occurred in Washington County as alleged by the state but, instead, was committed in Clatsop County when petitioner threw HF off the bridge. It followed, defense counsel argued, that the state had failed to prove the venue element of the offense. Defense counsel acknowledged the terms of ORS 131.325, but argued that that statute did not authorize venue in Washington County because the place of the victim’s death readily could be determined to be Clatsop County. As noted, counsel did not present any ground for the jury to acquit petitioner other than the venue defense, nor did counsel argue that there was any basis to convict peti- tioner of a lesser crime, although the jury was instructed on lesser-included offenses. The jury convicted petitioner of aggravated murder and sentenced him to death. B.  Evidence Adduced at the Post-Conviction Trial In this proceeding, petitioner asserted, in pertinent part, that both Hartshorne and Ferris were wrong about 694 Johnson v. Premo the cause of HF’s death, and that she actually died of a drug overdose. Petitioner alleged that his trial counsel was inad- equate in failing to develop and present that theory of death to the jury. More specifically, petitioner asserted that his counsel should have developed a theory that HF had con- sensually taken drugs and had sex with him,4 and that she had died of an accidental overdose. In support of that claim, petitioner adduced evi- dence that, early in the investigative process, he had told his defense team that, shortly before her death, HF had will- ingly consumed alcohol, marijuana, morphine and other opi- oids with him. He also indicated that they had engaged in sexual intercourse. According to petitioner’s account, he fell asleep, awakened at one point to find HF choking and vom- iting, and helped her to the bathroom. He fell asleep again, and when he awoke later, he discovered that she was dead. He then wrapped her in a blanket, put her in his car, drove her to the coast, then threw her off the Astoria Bridge. Conflicting evidence was presented as to whether petitioner also told the defense team that he was not sure whether HF was dead when he threw her off the bridge. The trial court did not explicitly resolve that factual issue. It did, however, make several findings of fact and conclusions of law that implicitly resolved that conflict in the evidence in favor of petitioner’s story that he had told counsel that the victim had died in Washington County.5 Accordingly, we accept the post-conviction court’s implicit factual determination that 4   Petitioner has referred at various points to the sexual contact as “consen- sual.” We understand petitioner’s use of the term “consensual” to be colloquial rather than legal. Under Oregon law, a 15-year-old cannot “consent” to sexual intercourse. See generally State v. Ofodrinwa, 353 Or 507, 300 P3d 154 (2013). 5   In particular, the post-conviction court stated: “Walker admits that Petitioner told him that Petitioner found HF dead in bed * * *. So trial counsel knew that unless Petitioner was quite mistaken over a period of some time and after close contact with the body handling it in different occurrences, it was likely as a fact that HF was dead when she left Washington County. * * * “* * * * * “[Walker and Peters] reasonably knew that even with difficulties of presenting morphine overdose (for example, it might reasonably ‘require’ Petitioner to take the stand), factually it was an option. * * * “* * * * * Cite as 361 Or 688 (2017) 695 defense counsel knew that the venue defense was inconsis- tent with their client’s version of the facts. The majority of petitioner’s evidence in support of the pertinent claim consisted of expert testimony that the victim did not die either from strangulation or from drown- ing, but, rather, that she died from a drug overdose. Among other evidence, petitioner presented testimony from Ferris, who, as noted, had testified at the criminal trial that the victim died from drowning. Ferris testified in the post- conviction proceeding that, despite his former opinion, he had come to believe that HF actually died of a morphine overdose. Ferris further testified that petitioner’s trial coun- sel had withheld information from him about petitioner’s account that he had found the victim dead in bed before throwing her off the bridge. The post-conviction court explicitly found Ferris not to be credible, concluding that his testimony was not “help- ful in any manner” to petitioner.6 However, the court did credit testimony that petitioner presented from two addi- tional expert witnesses, Dr. Julien and Dr. Ophoven. Julien, a retired anesthesiologist, opined that the amount of mor- phine in HF’s body was sufficient to have caused her death, and concluded that it was “very likely” that she had died of a morphine overdose, assuming she had not developed a tol- erance to opioids. Ophoven, a forensic pathologist, testified similarly that HF had a potentially lethal level of morphine in her bloodstream. Ophoven also testified that she did not observe clinical findings typically associated with drown- ing, and that the injuries to HF’s neck were not indicative of death by strangulation, given the lack of damage to internal structures in her neck. Although Ophoven could not defini- tively rule out other possible causes of death, she ultimately opined that HF likely had died of a morphine overdose. Like “[Walker and Peters] knew that if Petitioner was correct as to where she died, that as a fact, HF did not die by drowning because she was already dead.” (Emphasis in original.) 6   The post-conviction court noted extensive evidence in the record that coun- sel had informed Ferris that petitioner had asserted that HF had overdosed on drugs, which contradicted Ferris’s testimony at the post-conviction trial. 696 Johnson v. Premo Julien, Ophoven based her opinion on an assumption that HF had not built up a tolerance to opioids. The state’s expert witness at the post-conviction trial, Dr. Rosenblum, an anesthesiologist, did not offer an opinion as to the likely cause of HF’s death. However, he conceded that the amount of morphine in HF’s system “could actually represent a lethal dose.” C.  The Post-Conviction Court’s Decision The post-conviction court acknowledged that there was evidence that Walker had considered the morphine- overdose theory as a defense.7 However, the court opined that, if defense counsel had “considered and pursued” that theory, they would have either (1) not presented Ferris’s opinion as to the cause of death and instead have prepared to use petitioner as a witness in support of a drug overdose theory of death; or (2) sought the opinion of a toxicologist to support an overdose theory of death. The post-conviction court acknowledged that the first alternative alone would not have been viable, in light of petitioner’s vulnerability to impeachment on cross-examination,8 as well as the strength of the medical examiner’s testimony. The post-conviction court’s analysis, therefore, rested on whether counsel should have pursued the sec- ond possibility. The court acknowledged that an overdose theory might, in fact, have required petitioner to take the stand, but the court nevertheless concluded that “factu- ally it was an option.” (Emphasis in original.) The court explained that the drowning theory advanced by Ferris not only was inconsistent with petitioner’s version of the facts, but “there was virtually little chance that any jury would acquit upon Petitioner’s defense which acknowledged that 7   In particular, the post-conviction court cited Walker’s deposition testimony that Ferris’s confidence about the cause of HF’s death factored into his decision to pursue the drowning theory rather than a morphine-overdose theory. 8   As noted above, the defense team knew from pretrial rulings that evidence would be admitted about petitioner’s prior drugging and sexual assaults of other teenaged victims. In addition, petitioner had a criminal history and other pend- ing criminal charges. Given those factors, as well as the facts that petitioner had related to the defense team about what had occurred, we understand the post- conviction court to have recognized that adequate counsel would have understood the significant risks of having petitioner testify. Cite as 361 Or 688 (2017) 697 Petitioner killed [HF].” The court similarly described the venue defense that counsel had pursued as “more than just distasteful,” and one on which a jury was “very unlikely” to acquit. The post-conviction court concluded that, in light of the patent weakness of the venue defense,9 trial counsel could “not simply hide behind the expert’s opinion because trial counsel knew there was another factual defense which was at least plausible even if it was difficult under the cir- cumstances to prove.” The court ultimately concluded: “The point here isn’t that trial counsel made an informed choice after due diligence toward being well informed on potential defenses, but that trial counsel limited the uni- verse of options from which to make an informed decision by choosing not to consider hiring a toxicologist. Thus, trial counsel made a choice of defense without due diligence toward being informed.” The court further concluded that petitioner was prejudiced by counsel’s failure to pursue the overdose the- ory; the court noted, in particular, that all the experts at the post-conviction trial had testified that the amount of mor- phine found in HF’s system was, at least possibly, a fatal level. The court therefore granted petitioner relief in the form of a new trial. D.  The Court of Appeals Opinion On appeal from the ensuing judgment, the Court of Appeals rejected the state’s assertion that the post- conviction court’s conclusion meant, in effect, that adequate counsel always must consult additional experts to seek opin- ions supporting a factual theory advocated by the client, even after multiple experts already had ruled out that theory. Johnson, 277 Or App at 236. Relying on this court’s decision in Lichau v. Baldwin, 333 Or 350, 361, 39 P3d 851 (2002), and the post-conviction court’s determination that counsel knew that a drug overdose theory of defense was “plausi- ble,” the Court of Appeals concluded that counsel’s decision to limit the investigation of that defense “was not based on a 9  Peters acknowledged in the post-conviction trial that he understood the venue defense to be very undesirable. 698 Johnson v. Premo reasonable evaluation of the likely costs and potential ben- efits to petitioner.” Johnson, 277 Or at 239 (internal quota- tion marks and citations omitted). In particular, the Court of Appeals noted that the post-conviction court had found that “a request to hire a toxicologist would most likely have been granted because there would not have been any time or cost constraints associated with the request.” Id. As the Court of Appeals saw the calculus, the costs of pursuing the investigation would have been minimal, “while the benefit of pursuing the overdose defense could have been immense.” Id. The Court of Appeals also rejected the state’s argu- ment that petitioner had not shown that he was prejudiced by counsel’s failure to further investigate and present a defense based on a drug overdose theory. The state had asserted that, because petitioner was not willing to testify, he was not prejudiced by counsel’s failure to pursue that theory, which the state viewed as dependent on petitioner’s testimony. The Court of Appeals rejected that argument for two reasons. First, the court noted, the record did not establish that petitioner was unwilling to testify under any circumstances, but only that he was not willing to testify if the venue defense was presented. Id. at 241. Second, the Court of Appeals concluded, if necessary, counsel could have presented the factual basis for the overdose theory without petitioner’s testimony. Id. The Court of Appeals did not view the drug over- dose theory as a complete defense of the sort that could have resulted in petitioner’s outright acquittal. Rather, the Court of Appeals stated: “We emphasize that petitioner was prejudiced by his attorneys’ lack of due diligence even though evidence that the victim died of a morphine overdose might simply mean that petitioner caused the victim’s death in a different way. Testimony from a toxicologist that the victim died from a morphine overdose would have allowed petitioner’s defense counsel to argue that petitioner was guilty of a different offense—such as manslaughter, ORS 163.118(1)(a); crim- inally negligent homicide, ORS 163.145(1); or felony mur- der not committed intentionally, ORS 163.115(1)(b)—that, unlike aggravated murder, does not implicate the death Cite as 361 Or 688 (2017) 699 penalty. Alternatively, the evidence could have allowed the jury, even if it convicted petitioner of aggravated murder, to find that petitioner did not ‘deliberately’ cause the vic- tim’s death, a finding that would result in a sentence of life imprisonment without the possibility of parole but not in a sentence of death.” Id. at 240-41 (footnote and citation omitted). Accordingly, the Court of Appeals upheld the judgment granting peti- tioner post-conviction relief. On review before this court, the state contends that the post-conviction court and the Court of Appeals erred in concluding that counsel provided inadequate assistance. The state posits that, as a matter of law, when a qualified defense expert who has been informed of the relevant facts opines that the evidence does not support a possible defense theory, counsel is not constitutionally inadequate for fail- ing to seek additional expert opinions to contradict such an opinion.10 II. ANALYSIS To be entitled to post-conviction relief based on inadequate assistance of counsel, a petitioner must show that counsel failed to exercise reasonable professional skill and judgment, and that the petitioner suffered prejudice as a result of counsel’s inadequacy. Trujillo v. Mass, 312 Or 431, 435, 822 P2d 703 (1991). To demonstrate prejudice, a peti- tioner must show that counsel’s failure had “a tendency to affect the result of his trial.” Lichau, 333 Or at 359. Although this court interprets and applies Article I, section 11, inde- pendently of Sixth Amendment jurisprudence concerning the effectiveness of counsel, we have “recognized that the standards for determining the adequacy of legal counsel under the state constitution are functionally equivalent to those for determining the effectiveness of counsel under the federal constitution.” Montez v. Czerniak, 355 Or 1, 6-7, 322 P3d 487, adh’d to as modified on recons, 355 Or 598, 330 P3d 595 (2014) (citing cases). As the Court stated in Strickland v. Washington, 466 US 668, 688, 104 S Ct 2052, 80 L Ed 2d 10   The state has not entirely abandoned its arguments concerning prejudice, but it has not separately briefed that issue before this court, instead relying on its arguments before the Court of Appeals with respect to that issue. 700 Johnson v. Premo 674 (1984), to demonstrate ineffective assistance of counsel, a petitioner must show that trial counsel’s performance “fell below an objective standard of reasonableness.” And, if a petitioner proves that counsel was ineffective, he or she also must show that there was a “reasonable probability that, but for counsel’s unprofessional errors, the result of the pro- ceeding would have been different.” Id. at 694. In evaluating counsel’s performance, we view the conduct in question “without the distorting effect of hind- sight.” Lichau, 330 Or at 360. That, however, can be eas- ier said than done. In every post-conviction case, a court is called on to evaluate counsel’s actions and omissions in circumstances that resulted in the petitioner’s conviction. Thus, in hindsight, a court may tend to scrutinize counsel’s decisions in an overly critical light, given the known out- come, and be more inclined to conclude that the petitioner received inadequate assistance of counsel. That type of hind- sight reflects “outcome” bias. See generally Stephanos Bibas, The Psychology of Hindsight and After-the-Fact Review of Ineffective Assistance of Counsel, 2004 Utah L Rev 1 (2004) (discussing this type of bias). In contrast, given the known outcome of conviction, there may be a tendency to evaluate the second prong of the inquiry—whether counsel’s ineffec- tive assistance had a tendency to affect the outcome—with a confirmation bias. That is, in hindsight, there may be a tendency to view counsel’s errors as having had no effect on what may seem to have been an inevitable or “foreordained outcome.” Id., 2004 Utah L Rev at 3. In sum, in the absence of disciplined scrutiny, the distorting lens of hindsight could make a court more likely to view counsel’s decisions as inad- equate, but make it less likely to view counsel’s errors as having had a tendency to affect the outcome. In the present case, as explained below, the ques- tion whether counsel provided inadequate assistance is a close one. But, unlike in many cases, this is not the type of case in which it would one would readily conclude that there was no prejudice. That is, unlike cases in which a ret- rospective view might suggest a foreordained outcome, in this situation, there is a significant dispute concerning one of the most crucial facts in any death penalty case: How did Cite as 361 Or 688 (2017) 701 the victim die? And even where, as in this case, there is no serious dispute that petitioner caused the victim’s death, a jury’s determination of how the death occurred not only could affect the degree of the petitioner’s culpability, but also obviously could affect the jury’s determination whether petitioner should receive the death penalty. Accordingly, our focus—like the state’s focus in its arguments on review—is primarily on the question whether counsel provided inade- quate assistance. A.  Ineffective Assistance of Counsel under Article I, section 11 As noted, to establish inadequate assistance of counsel, a petitioner must demonstrate by a preponderance of the evidence that counsel “failed to exercise reasonable professional skill and judgment.” Montez, 355 Or at 7. As we have long recognized, “any statement of the standard of per- formance constitutionally required of counsel must neces- sarily be general.” Krummacher v. Gierloff, 290 Or 867, 873, 627 P2d 458 (1981). That is, what constitutes adequacy in formulating and executing a defense depends on the “nature and complexity of the case.” Id. at 875. As for the “nature” of the underlying criminal case here, petitioner was charged with aggravated murder, and the state sought the death penalty. Simply put, there is no greater crime, or more severe penalty, under state law, and no type of criminal case requires more care in preparation. In addition, the factual complexity of petitioner’s crimi- nal case was manifest. It is true that part of the evidence was straightforward: The state had strong evidence that petitioner, who had a history of drugging and raping teen- aged girls, drugged and raped HF and then killed her. In addition, counsel knew that petitioner had acknowledged that he provided the drugs to the victim, that he had sex with her, and that he threw her off a bridge. However— and taking care not to view the issue in hindsight—the forensic evidence known to the defense team was far from conclusive. The information provided by the state’s medical examiner and the defense pathologist was extensive, com- plex, and in conflict as to the cause of the victim’s death. Both experts recognized, in developing their opinions, that 702 Johnson v. Premo there was much data to be considered: There were physi- cal injuries to the victim’s body; water, silt and aspirated vomit in her lungs; semen in her vagina; and drugs in her bloodstream. In sum, there was overwhelming evidence that numerous crimes had been committed against the 15-year- old victim. That evidence, however, provided more questions than answers as to precisely how she had died. This was a complex case in terms of the forensic evidence. And, it was a death penalty case. Those circumstances counseled care in the development of a trial strategy that, in the best of circumstances, would be difficult. Of course, the test for adequacy of assistance of counsel “allows for tactical choices that backfire, because, by their nature, trials often involve risk.” Krummacher, 290 Or at 876. “[I]f counsel exercises reasonable professional skill and judgment, a reviewing court will not second-guess the lawyer in the name of the constitution, but neither will the court ignore decisions made in the conduct of the defense which reflect an absence or suspension of professional skill and judgment.” Id. at 875-76. Before assessing whether a tactical choice is reasonable, though, there is a preliminary factual question as to whether a choice was made at all. See Pereida-Alba v. Coursey, 356 Or 654, 670-73, 342 P3d 70 (2015) (remanding for post-conviction court to make fac- tual determination, given conflicting evidence in record as to whether counsel made conscious choice about whether to forego requesting jury instruction); Green v. Franke, 357 Or 301, 320-21, 350 P3d 188 (2015) (remanding for determina- tion whether counsel made tactical choice not to request lim- iting instruction). As this court noted in Pereida-Alba, “whether the failure to consider an issue constitutes inade- quate assistance will turn on, among other things, whether the strategy that defense counsel did employ was rea- sonable, the relationship between the evidence or theory that defense counsel failed to consider and the strategy that counsel did pursue, and the extent to which counsel should have been aware of the strategy that petitioner now identifies.” Pereida-Alba, 356 Or at 674 (citing Montez, 355 Or at 24). Cite as 361 Or 688 (2017) 703 In this case, the post-conviction court determined that “trial counsel limited the universe of options from which to make an informed decision by choosing not to con- sider hiring a toxicologist. Thus, trial counsel made a choice of defense without due diligence toward being informed.” In light of the broader context of the court’s decision, we under- stand it to have meant that trial counsel did not make a reasonable tactical decision not to pursue a drug overdose theory: There was uncontradicted evidence that Walker considered the theory and rejected it in light of Ferris’s cer- tainty that HF had drowned and the potential problems with having petitioner testify. Thus, we understand the post-conviction court to have determined that counsel made a tactical decision to forego a morphine-overdose defense in favor of a venue defense, but that that decision was unrea- sonable because counsel had not undertaken an adequate investigation of the relevant facts that, in turn, would have required hiring a toxicologist to review the forensic evidence. As this court has observed on numerous occasions, “tactical decisions must be grounded on a reasonable inves- tigation.” Gorham v. Thompson, 332 Or 560, 567, 34 P3d 161 (2001) (citing Krummacher, 290 Or at 875; Stevens v. State of Oregon, 322 Or 101, 108, 902 P2d 1137 (1995)). In Lichau, this court considered whether the petitioner’s trial counsel had made a “reasonable investigation” in the context of his decision to withdraw from the jury’s consideration the peti- tioner’s alibi defense. We discuss that decision, and Stevens, in some detail, because those decisions provide helpful guid- ance in determining what constitutes a reasonable investi- gation of a factual theory of defense. In Lichau, the petitioner had been charged with numerous sexual offenses against his 11-year-old niece. 333 Or at 352. At the time of the alleged offenses, the petitioner had been stationed at a military base on the East Coast. Defense counsel prepared an alibi defense, based on mili- tary records, that the petitioner was not in Oregon when the offenses were committed. On the morning of trial, how- ever, the prosecutor threatened to present evidence that did not appear on official military leave records that the peti- tioner had been on military leave in Oregon at the time of 704 Johnson v. Premo the crimes. Id. at 353. Counsel withdrew the alibi defense in anticipation of reviewing the prosecutor’s evidence, but the prosecutor never provided the evidence. Id. In the ensuing post-conviction action, the petitioner argued that adequate counsel would have determined, after a sufficient investigation, that the petitioner had not, in fact, been on leave in Oregon at the time of the offenses or, alternatively, would have reinstated the alibi defense when the prosecutor failed to provide the touted evidence about petitioner’s military leave. Id. at 354. Evidence at the post- conviction hearing established that defense counsel had relied entirely on military records provided by the prosecu- tion, and had not attempted to subpoena any military records before trial. Id. Evidence at the post-conviction hearing also showed that the petitioner had not been on leave at the time of the alleged offenses, and that proof of that fact could have been obtained if counsel had subpoenaed the petitioner’s military records, interviewed witnesses, and obtained other documents such as the petitioner’s bank statements and cor- respondence. Id. at 356-57. Ultimately, this court agreed with the post- conviction court that defense counsel had failed to conduct an adequate investigation, noting that counsel had limited his preparation of the alibi defense to reviewing materials pro- vided by the state, seeking information from the petitioner’s prior lawyer (who provided no relevant information), and contacting the petitioner’s parents. Id. at 360-61. This court acknowledged that the petitioner had told defense counsel that he was uncertain of his whereabouts at the time of the alleged crime, but, in this court’s view, that fact was not dis- positive: counsel knew that the petitioner had been stationed at a military base thousands of miles away, and the peti- tioner had told counsel “that military records demonstrating that he had no taken leave in June 1989 were available and that supervisors might be able to confirm” that he was at the base at the pertinent time. Id. at 361. This court therefore concluded that counsel’s decision to limit his investigation of the alibi defense was not “based on a reasonable evaluation of the likely costs and potential benefits” to the petitioner. Id. (quoting Stevens, 322 Or at 109). Cite as 361 Or 688 (2017) 705 Stevens likewise involved a failure to adequately investigate a possible factual theory of defense. In that case, the petitioner was charged with the rape of a child that was alleged to have occurred as he drove her to school. The child testified to the details of the alleged offense, and further testified that upon arrival at school, she had told one of her teachers what had occurred. No physical evidence corrobo- rated the crime. Id. at 103-04. Although the petitioner told defense counsel that he was not physically capable of having committed the crime due to impotence, counsel did not seek information about the petitioner’s physical condition from his urologist. Id. at 104, 106. In addition, counsel did not interview the child’s teachers or peers at her school. Id. at 105. Had counsel done so, he would have discovered that the child had not, in fact, reported a rape to her teacher, that she had told various classmates different stories about what had occurred, and that she had revealed to classmates that her mother intended to sue the petitioner for “money.” Id. at 105-06. This court stated: “In investigating a case, a lawyer inevitably is faced with choices as to what avenues of investigation to pursue. A ‘tactical decision’ in the course of an investigation is a conscious choice by a lawyer either to take or to omit some action on the basis of an evaluation of the nature and com- plexity of the case, the likely costs and potential benefits of the contemplated action, and other factors. But the fact that a lawyer has made a ‘tactical decision’ does not mean that the lawyer’s choice meets the constitutional standard for adequate assistance of counsel. Indeed, this case illustrates the point. Considering the nature and complexity of this case, trial counsel’s choice not to interview the complaining witness’s teachers and classmates was a ‘tactical decision,’ but it did not result in adequate representation of peti- tioner. The complaining witness’s statements to classmates suggested a possible motive to fabricate (i.e., obtain money through litigation). The teachers’ testimony would have impeached the testimony of the parents of the complaining witness. The conflicting accounts to classmates regarding the location of the alleged rape would have impeached the complaining witness’s account at trial. Medical evidence of impotence would have contradicted part of the complaining witness’s testimony. 706 Johnson v. Premo “Trial counsel’s decision not to interview potential wit- nesses at the complaining witness’s school was not a choice that was based on a reasonable evaluation of the likely costs and potential benefits of pursuing the investigation.” Id. at 109. In some ways, the present case was more compli- cated factually than either Lichau or Stevens. Here, there was no evidence that counsel knew of some specific witness or evidence that might assist in developing a drug overdose theory, yet failed to follow through with contacting that per- son or seeking that evidence. Still, counsel had a wealth of relevant information to consider. They knew that their cli- ent believed that the victim had died in his bed after he had given her drugs and had sex with her, and they at least implicitly understood that petitioner believed that the vic- tim had died of a drug overdose. They knew, in contrast, that the medical examiner who examined the victim’s body believed that she had been strangled. Given the discrepancy between what petitioner said had occurred and the medical examiner’s conclusion, coun- sel needed to—and did—obtain more information about the cause of the victim’s death. In particular, counsel obtained the opinion of an experienced forensic pathologist who had performed thousands of autopsies. In the normal course, counsel might reasonably expect such an expert either to confirm the client’s version of what had occurred or confirm the state’s version of events. This case, however, did not fol- low that course. Instead, the expert added a third possible theory of how the victim’s death had occurred: by drowning. Contrary to petitioner’s suggestions, it is unlikely that any of the three possible theories of death would have produced a strong defense in his criminal case. Bluntly, if the theory was death by strangulation, there was ample evidence that petitioner strangled the victim; if the theory was death by drowning, there was ample evidence that peti- tioner drowned the victim, albeit not in the county in which the prosecution was commenced; and, if the theory was death by drug overdose, there was ample evidence that peti- tioner gave the victim the drugs that caused the overdose. Moreover, although the state pursued the strangulation Cite as 361 Or 688 (2017) 707 theory, none of the three theories of death necessarily was incompatible with a charge of aggravated murder committed in the course of, or to conceal, sex crimes committed against the victim. Although petitioner suggests that pursuing the overdose theory could have led to an outright acquittal on the charges, that seems at least as improbable as the pros- pects for the venue defense that was presented. Petitioner’s position appears to be that a factual defense could have been presented that HF had voluntarily ingested the drugs, vol- untarily had sex with petitioner, and then accidentally died from an overdose. But, such a theory would have required counsel to argue that HF was neither physically nor men- tally incapacitated due to the allegedly lethal dose of mor- phine when petitioner had sex with her, and that the sexual contact was not forcible. As noted, the state presented evi- dence to the contrary,11 and convincing a jury that the victim was not incapacitated by an amount of morphine that ulti- mately killed her could have proven extremely difficult, even if petitioner had testified to that effect in his own defense. Nonetheless, pursuing such a defense had one major advantage—unlike the competing cause of death the- ories, neither of which could easily be characterized as unin- tentional, with a drug overdose theory, counsel could have argued that the homicide did not constitute aggravated mur- der because it was not committed intentionally.12 Although such an approach would have left petitioner vulnerable to conviction for a variety of lesser offenses, it had the potential to remove the death penalty from the equation. Simply put, it is impossible to overstate the importance of that consider- ation in a capital murder case. In addition, even if an overdose theory had not succeeded in the guilt phase and the jury had convicted of aggravated murder, that theory would have laid better 11   As this court observed in the opinion on direct review, the state presented evidence that HF was a lesbian in support of its theory that she would not have willingly had sexual contact with petitioner. Johnson, 340 Or at 342. 12   Every charge of aggravated murder that the state alleged in the underly- ing criminal proceeding was based on allegations that the crime was committed intentionally. 708 Johnson v. Premo groundwork for arguing in the penalty phase that the jury should not impose a sentence of death. In the penalty phase, the jury was required to answer four questions, including a specific question about whether the offense was commit- ted “deliberately,” and a more open-ended question about whether “the defendant should receive a death sentence.” ORS 163.150(1)(b)(A), (D). It is reasonably possible that a jury, in considering those questions, could view a death caused by drug overdose as less heinous than one caused by strangulation or intentional drowning, and therefore be less inclined to impose the death penalty. The state counters that petitioner’s “refusal to coop- erate [with defense counsel] left them with little confidence in any approach based on a theory of accidental death.” (Emphasis added.) That, however, is an overstatement. It is true that the record shows that petitioner had no interest in pursuing a strategy that would have involved acknowl- edging that he had strangled the victim. It is also true that petitioner was not forthcoming with his defense team and was unhelpful in numerous respects. Those facts, however, do not foreclose the possibility that petitioner would have been more cooperative if counsel had pursued a strategy based on facts that matched, at least in a general sense, the account that petitioner gave his defense team about the cir- cumstances of the victim’s death. The state remonstrates that the field of forensic pathology includes interpretation of toxicological tests, and that both of the criminal trial expert’s reports included such interpretations. Thus, the state suggests, adequate defense counsel should not have been expected to seek additional information from a toxicologist because they already had expert opinions that encompassed the subject of toxicology. To be sure, the state’s position has some force. We do not suggest that, whenever a homicide involves the use of drugs, the opinions of forensic pathologists about the cause of the victim’s death must be supplemented by information from a toxicologist. Here, however, the information that the defense team had at hand should have prompted adequate counsel to seek additional information about the effects of the drugs found in the victim’s body. First, as noted, petitioner told counsel that he and the victim had taken not only morphine, Cite as 361 Or 688 (2017) 709 but numerous other drugs, and also had consumed alcohol. He also told the defense team that the victim had vomited after taking morphine. Second, although the medical exam- iner ruled out morphine overdose as a cause of death, his report indicated that her morphine level was significant enough that HF would have had difficulty resisting an assault. He also indicated that there was evidence that HF had aspirated vomit at some point before her death and that a drug screen had revealed marijuana metabolites in HF’s urine. Third, although Ferris, like the medical examiner, ruled out morphine overdose as a cause of death, his report suggested that the victim likely had aspirated vomit while unconscious due to morphine, and he also opined that mor- phine levels are very difficult to interpret. Notably, the forensic data available to defense coun- sel tended to confirm important aspects of petitioner’s ver- sion of what had occurred—that HF may have had multi- ple intoxicating substances in her system at the time of her death, and that she had vomited due to morphine ingestion at some time before her death. In sum, all the evidence indi- cated that drugs had played a significant role in the events that surrounded HF’s death. As the state points out, when counsel has “reason to believe that pursuing certain inves- tigations would be fruitless or even harmful, counsel’s fail- ure to pursue those investigations may not later be chal- lenged as unreasonable.” Strickland, 466 US at 691. Here, however, the conflicting information from the experts, not only about the cause of HF’s death but also about the role that morphine and other intoxicants may have played in the circumstances surrounding her death, was sufficient to give adequate counsel reason to believe that further inquiry about toxicology would not necessarily be “fruitless.” As for the extent to which it could have been “harmful,” counsel was well aware before trial commenced that the state would present evidence of petitioner’s history of drugging teen- aged girls with morphine in order to have sexual contact with them. In other words, counsel knew that “harmful” evidence would be presented about the victim having poten- tially incapacitating levels of morphine in her body, as well about petitioner having likely given her that morphine. In those circumstances, seeking additional information about 710 Johnson v. Premo the significance of the morphine and other drugs in HF’s system should not have been foregone on the premise that such a quest was likely to be “harmful” to the defense. What constitutes “reasonable professional skill and judgment” in defending criminal charges is a highly fact-specific inquiry. Here, the information that defense counsel had—both from their own client and from discov- ery and independent investigation—suggested no defense that had great merit. As a matter of common sense, though, counsel not only should have realized—but did realize—that a venue defense provided no reasonable prospect for acquit- tal. Moreover, that defense had the significant drawback of essentially acknowledging that petitioner had committed aggravated murder, and had done so in a particularly cal- lous manner by throwing a youth whom he had sexually assaulted off a bridge. In contrast to that very weak strategy, petitioner adduced expert testimony in the post-conviction trial that, in a death penalty case, adequate counsel should attempt, whenever possible, to develop a unified defense theory for both the guilt and penalty phases of the trial, with an eye toward minimizing the risk that a jury that convicts will impose the death penalty. The defense advanced in this case, at best marginally viable in the guilt phase, lacked any tactical value in the penalty phase. In that admittedly very difficult circumstance, and in light of all the informa- tion known to defense counsel as described above, we con- clude that adequate trial counsel should have sought out additional information concerning the drugs in the victim’s system at the time of her death, in order to try to develop a guilt-phase theory that petitioner’s killing of the victim was unintentional, or alternatively, a penalty-phase theory that, even if the killing was intentional, it was not the type of crime for which the death penalty should be imposed. B. Prejudice As noted, to obtain post-conviction relief, a peti- tioner must show that counsel’s inadequacy had “a tendency to affect the result of his trial.” Lichau, 333 Or at 359. And, as noted, the state makes no specific argument on review concerning prejudice. As our analysis indicates, the theory Cite as 361 Or 688 (2017) 711 of defense advanced by counsel in this case had little chance of success at the guilt phase of the trial, and it left even less room in the penalty phase for counsel to argue that the cir- cumstances of the victim’s death did not justify imposition of the death penalty. On the other hand—although unlikely to produce an acquittal—pursuing a drug-overdose defense held better promise, especially in the penalty phase of this aggravated murder case. We therefore conclude that peti- tioner has demonstrated that counsel’s failure to adequately investigate that defense had a tendency to affect the result of his trial. The decision of the Court of Appeals and the judg- ment of the circuit court are affirmed.