Paul M. McManus v. Ron Neal

779 F.3d 634, 2015 U.S. App. LEXIS 2389, 2015 WL 667466

• Court: Court of Appeals for the Seventh Circuit
• Decided: February 17, 2015
• Docket: 12-2001
• Status: Published

Opinion

SYKES, Circuit Judge.

An Indiana jury convicted Paul Mc-Manus of murdering his estranged wife and two young daughters, and the trial judge sentenced him to death in accordance with the jury’s recommendation. The Indiana Supreme Court affirmed on direct appeal, but on postconviction review the trial judge found McManus intellectually disabled and thus ineligible for the death penalty. See Atkins v. Virginia, 536 U.S. 304, 122 S.Ct. 2242, 153 L.Ed.2d 335 (2002); see also Ind.Code § 35-36-9-6. A divided Indiana Supreme Court disagreed and reimposed the death sentence.

McManus then sought federal habeas review on several claims of constitutional error, including a challenge to the rejection of his claim of intellectual disability under Atkins. The district court denied relief but authorized an appeal on the Atkins issue. We expanded the certificate of appealability to include the following questions: (1) whether the state courts unreasonably applied federal due-process standards in finding McManus competent to stand trial, see Pate v. Robinson, 383 U.S. 375, 86 S.Ct. 836, 15 L.Ed.2d 815 (1966); Dusky v. United States, 362 U.S. 402, 80 S.Ct. 788, 4 L.Ed.2d 824 (1960); (2) whether McManus was forced to appear before the jury in a “drug-induced stupor” in violation of Riggins v. Nevada, 504 U.S. 127, 112 S.Ct. 1810, 118 L.Ed.2d 479 (1992); and (3) whether McManus’s trial attorneys were ineffective for failing to present additional mitigating evidence about his intellectual disability during the sentencing phase of the trial.

We agree with the district court that McManus is not entitled to habeas relief on his claim of categorical ineligibility for the death penalty. The state high court applied the rule of Atkins and made a reasonable factual determination that Mc-Manus is not intellectually disabled. But the state courts unreasonably applied clearly established due-process standards for adjudicating a defendant’s competency to stand trial. The record reflects that McManus decompensated soon after the trial testimony got underway. He had several panic attacks, and his symptoms were severe enough to require two trips to the emergency room. There he was treated with a potent combination of several psychotropic drugs—including one that knocks out memory—as well as an opioid painkiller. He remained on a regimen of mind-altering medications for the duration of the trial.

The powerful effect of the medications alone created substantial doubt about Mc-Manus’s mental fitness for trial, but the judge never ordered a competency evaluation. Instead, the judge focused on getting McManus “fixed up” enough to complete the trial. By taking this approach, the judge failed to apply the legal framework established in Dusky and Pate for addressing competency questions. The Indiana Supreme Court recited the correct legal standard but in the end did not actually apply it. Although habeas review of state judgments is deferential, see 28 U.S.C. § 2254(d)(1)—(2) (2012), the record does not permit a conclusion that the state courts reasonably applied federal constitutional requirements for adjudicating a defendant’s competency to stand trial.

Accordingly, we reverse and remand to the district court with instructions to grant the writ unless Indiana gives notice of its intent to retry McManus within a reasonable time to be set by the district court. This holding makes it unnecessary for us to address McManus’s remaining claims, which rest on other allegations of constitutional error at trial.

I. Background

A. The Murders, Trial, and Posttrial Motion to Correct Errors

Habeas review in capital cases usually entails a lengthy procedural record, and this case is no exception. We limit our historical account of the case to the details that are important to the claims on which the appeal was authorized. Even so, significant length cannot be avoided.

Paul McManus married his wife, Melissa, in 1992. They had two daughters, Lindsey and Shelby, and the family lived in Evansville, Indiana. Shelby, the younger girl, had serious birth defects. She was bom without eyes and her esophagus did not connect to her stomach; she received nourishment through a feeding tube.

At the time of the crimes, McManus was working three jobs: He was a laborer at a plastics factory, a barback at a local pool hall (he stocked the bar with ice and beverages and otherwise assisted the bartender), and one day a week he did janitorial work at a freight company.

In the fall of 2000, Melissa left Paul, taking their daughters with her. At the time Lindsey was almost eight years old and Shelby was not quite two. The couple officially separated in December, although Melissa and the girls continued to live in Evansville.

On January 24, 2001, McManus was arrested for domestic battery against his estranged wife. Melissa told the arresting officer that McManus had threatened to kill “everyone.” During the next few weeks, McManus talked of suicide and continued to threaten violence against his family. He was fearful that Melissa would leave Evansville with the girls, and he spoke of wanting to kill himself and his family so they could be together.

On the morning of February 26, 2001, McManus was served with divorce papers. Later that day he carried out his threats against his family. He got a handgun from his brother’s house, bought ammunition at a gun store, and took a taxi to his wife’s home. There he shot Melissa once in the leg and three times in the head. Turning the gun on the girls, he shot Lindsey three times in the head and Shelby once, also in the head. After killing his family, McManus took Melissa’s car, left the scene, and called his mother and sister to confess what he had done. Then he drove to the Ohio River Bridge, climbed to the top, and threw himself into the river. Law-enforcement officers saw the jump and rescued him.

McManus was charged with three counts of murder. Indiana sought the death penalty, citing the multiple murders and the murder of two persons under the age of 12 as statutory aggravating factors. See Ind.Code § 35-50-2-9(b)(8), (12) (2013). McManus’s counsel filed a notice of intent to assert an insanity defense, so the judge postponed the trial to accommodate the forensic psychiatric examinations required to mount that defense.

For 14 months while in pretrial detention, McManus was treated with the antidepressant drug Elavil and also a beta-blocker to control his anxiety. Trial was scheduled for April 24, 2002. About a month before trial, the jail psychiatrist changed McManus’s medication regimen, tapering his doses of Elavil from March 25, 2002 until April 17, 2002, then eliminating that drug altogether and substituting Effexor, another antidepressant. Expert testimony later established that Effexor can aggravate anxiety in some patients. Jail medical personnel also discontinued McManus’s beta-blocker, apparently out of a concern that it was exacerbating his depression.

Voir dire began as scheduled on April 24. By April 29 a jury was sworn and testimony began. During the noon recess on the first day of testimony, McManus suffered a panic attack. He was hyperventilating, his blood pressure was elevated, and he reported chest pain. His symptoms were severe enough that he had to be taken to the hospital, so the judge recessed the proceedings for the remainder of the day. McManus was treated in the emergency room and returned to the jail.

The next day McManus had another panic attack, with the same symptoms as the day before. His attorneys reported having great difficulty communicating with him and advised the court that he was not competent to assist the defense or decide whether to testify. The judge again recessed the proceedings and sent McManus back to the hospital. This time the judge called ahead and spoke to Dr. Reza Mo-hammadi, one of the emergency-room physicians, apparently to let him know that McManus was coming, although the record does not reflect exactly what was said during the phone call.

Dr. Mohammadi treated McManus with several intravenous medications: Versed (a drug used to treat seizures and to achieve sedation and amnesia during medical procedures); morphine (an opioid for pain); and Xanax (a psychoactive drug used to treat panic and anxiety disorders). Before releasing McManus back to the jail, Dr. Mohammadi prescribed oral Xanax and Lortab, a combination of acetaminophen and hydrocodone, an opioid. The Xanax prescription specified a dosing regimen of three times per day—down from the usual four—because the drug has a sedative effect.

Before resuming the trial, the judge summoned Dr. Mohammadi to the courtroom to question him about McManus’s condition. McManus was not in the courtroom during this testimony. The judge asked Dr. Mohammadi if the drugs he had given McManus were “mind altering” or would “affect a person’s mental processes.” The doctor replied “Absolutely.” He explained that “if the medicine is given to someone who’s not having any problems like this gentleman, it would probably put you to sleep and you will not be able to interact, period.” But “when someone is as anxious as this gentleman was, it probably would bring him down to a level that he can actually communicate.” Dr. Mo-hammadi cautioned, however, that patients who are treated with “this type of medication” are routinely instructed not to drive for four to six hours because “we believe it does alter their decision making and so on and so forth.”

The judge pressed the doctor to elaborate:

Q: But the medications you gave him today, would it prevent him from thinking rationally?

A: I would say that he would—it would alter the way he would perceive things. Now, in the spectrum of what we are dealing with today, I would say that he would be thinking more rationally now than he was when he was so anxious, if that answers your question.

Q: Yes. And how about the medications that you prescribed for him, the Xanax and the Lortab, how would they affect his mind and his judgment?

A: I believe he can-—he can make judgments in—if he was given enough time to make the judgment at, and again, it’s a decision that if, in fact, this man is not—if his condition is not controlled, he would not be able—in the state of mind he presented today, he would not be able to answer any questions rationally, period, and now that he’s on medicine, he may be—in my view, he can possibly now proceed and give some rational answers, but these medicines do alter—alter people’s judgment in the vast majority of people, yes they do.

The prosecutor asked the doctor if Mc-Manus would be able to recognize his attorneys and understand that they were “trying to help him be found not guilty.” Dr. Mohammadi replied, “I believe so.” He also said that McManus should be able to follow the trial testimony for at least the next few hours. 1 But his testimony was equivocal; the doctor cautioned that “[i]t’s very difficult on one encounter in an emergency room to decide what a patient’s response to a medicine would be.” And he qualified his testimony even further based on the limited scope of his expertise: he was an emergency-room physician, not a psychiatrist. In response to questions from defense counsel, Dr. Mohammadi could not predict how McManus would respond to the prescribed oral medication or whether his condition was likely to improve. He also acknowledged that Mc-Manus would need to be seen by a psychiatrist to determine what medication was appropriate to treat his symptoms yet permit him to understand and participate in the trial.

The uncertainties in Dr. Mohammadi’s testimony prompted the prosecutor to suggest that the doctor talk to McManus in the holding cell to get “a better feeling for how well he can respond or how well he’s doing on the medication.” The judge agreed and invited the doctor to “go back and talk to him and see what you think.”

Dr. Mohammadi talked to McManus in the holding cell and reported back that he was “more calm” and able to answer a few basic questions but had difficulty with others. For example, McManus knew what year it was, but he was unsure about the month and day. The judge asked the doctor if he found McManus to be “rational right now.” Dr. Mohammadi replied, “[rjight now, he’s definitely rational.” After a few more questions from the court and counsel, Dr. Mohammadi was excused.

McManus’s attorneys moved for a mistrial or, alternatively, a continuance so that McManus could be examined by a psychiatrist for competency to stand trial and stabilized on appropriate medication. The judge summarily denied the motion:

THE COURT: ... I believe that the defendant is competent to assist in his own defense and I’m not convinced that the situation would improve any more over the next few weeks than it is right now and that the—if he stays on his medication, that he will be rational— remain rational and be able to assist in his defense and understand the proceedings against him, so I’m going to deny the motion for a continuance and/or mistrial.

Trial resumed. When McManus was escorted into the courtroom, however, he could not walk under his own power. His lawyer noted for the record that McManus “had to be helped in by the sheriff’ and again asked the judge for a continuance and a psychiatric examination. The judge acknowledged that McManus “doesn’t appear to be in the condition the doctor testified he was in.” Nonetheless, the judge denied the motion without further comment and called on the prosecutor to present his next witness.

The following day—May 1, the third day of testimony—McManus' again became ill, complaining of light-headedness and nausea. His counsel reported that McManus was sick and renewed the mistrial motion; the judge again denied it and pressed on with the trial. Later that day McManus had another panic attack. He was hyperventilating and said he felt like the room was getting smaller. His counsel alerted the court that McManus was “about to fall out of his chair” and seemed like he was going to faint. The judge called another recess while McManus was treated in the jail infirmary. During the recess, the judge put one of the courtroom deputies under oath to make a record about Mc-Manus’s condition. The deputy testified that the jail nursing staff was tending to McManus but he was “still having a hard time getting his breath. He’s hyperventilating. They cannot get his breath calmed down, so they’re going to give him a shot of what, I do not know, but the nurse said it will knock him out for hours.”

Defense counsel again moved for a mistrial. The judge deferred ruling but ordered a one-week continuance, excusing the jury until Wednesday, May 8. The purpose of the continuance, however, was not to have McManus examined by a qualified expert for an opinion about his competency to stand trial. Instead, the judge intended to meet with jail medical personnel, “get ahold of a psychiatrist,” and “have sort of a confab so we’re all on the same page of what needs to be done” to get McManus “fixed up” enough to proceed. The judge made it clear that he intended to move forward with the trial: “[W]e need to do it as soon as we can to get him fixed up—whatever it takes to get him fixed up for next Wednesday.”

The judge contacted Dr. Willard Whitehead, a psychiatrist at the Southwestern Indiana Mental Health Center, and asked him to examine McManus in the jail. No record was made of the judge’s instructions to Dr. Whitehead, but it’s clear from the doctor’s report and testimony that he was brought in to consult on McManus’s treatment, not to conduct a competency examination.

On May 6 McManus’s defense team filed a verified motion for a mistrial, explaining that McManus had suffered two more panic attacks during the continuance and was unaware of what had occurred thus far during the trial. They also asserted that McManus lacked the ability to communicate with counsel or assist in his own defense, and was in no condition to make critical trial decisions such as whether to testify.

Dr. Whitehead saw McManus on May 2 and again on May 7, just before the hearing on the latest mistrial motion. The May 2 consultation is memorialized in a written report, but the visit on May 7 was apparently quite brief; no report is in the record.

To prepare for the May 2 meeting, Dr. Whitehead reviewed McManus’s jail medical records, but he did not read the reports of the mental-health experts who had examined McManus for purposes of the insanity defense. At the beginning of the interview, Dr. Whitehead told McManus that he was not evaluating him forensically but instead was there to help him feel better. The doctor thereafter had difficulty obtaining a psychiatric history from Mc-Manus and ultimately could not complete the examination. Although McManus was “pleasant and cooperative” and showed no signs of medication intoxication, he spoke and moved slowly and had trouble understanding directions. He was able to answer some initial questions about the symptoms he experienced during the panic attacks. Dr. Whitehead catalogued them as follows: shortness of breath, a racing heart, “needle-like pains in his head,” “heavy pain in his chest,” nausea, feeling hot or cold, “feeling unreal,” and not being able to feel his arms or face. Dr. Whitehead’s initial impression was that the attacks were caused by the stress of the trial and not an underlying panic disorder. After these initial questions, however, the interview was cut short because McManus experienced another attack: he began hyperventilating and was lying on the floor, unable to talk any further.

Because the examination could not be completed, Dr. Whitehead’s observations about McManus were necessarily tentative and qualified. He explained that “[t]here were some aspects of the interview that I didn’t get to even start because of that attack.” And his brief visit with McManus just before the hearing didn’t add much to his font of knowledge about his patient; the doctor said he found it “very hard to collect meaningful information today.” Dr. Whitehead also said he could not form an opinion about whether the panic attack was faked or self-induced or whether Mc-Manus was malingering.

Despite the limits on his examination, Dr. Whitehead did order a change in Mc-Manus’s medication. He removed Effexor, apparently because it can exacerbate anxiety, and he substituted Remeron, another antidepressant. He also put McManus back on a beta-blocker to try to achieve better control over his anxiety. Mc-Manus’s other medications—most notably Xanax—were continued. With these adjustments, Dr. Whitehead advised the court that McManus was receiving appropriate treatment, although he acknowledged that achieving the right balance was “a little bit of a tightrope between intoxicating and undertreating.” Dr. Whitehead was unable to testify to the precise effect of Dr. Mohammadi’s treatment—in particular, his use of Versed to calm McManus’s panic attack. He said he was not well-acquainted with that drug, although he understood that it “knocks out memory. I think that’s one reason they use it.” And he agreed that combining that medication with morphine would significantly slow a person’s mental acuity.

Dr. Whitehead was not asked to state an opinion about McManus’s competency to stand trial under the standard established in Dusky. He did not independently offer such an opinion.

At the end of the hearing, defense counsel again asked the court to order a mistrial because McManus was incompetent to proceed. In a brief bench ruling, the judge denied the motion:

THE COURT: Okay. I’m going to deny the motion. I’m convinced that it’s either self-induced, or if not self-induced, it’s something that’s caused by this trial. I think these—this—these doctors are giving him the optimum treatment he can get. I’m convinced that we’re not going to face any better situation the next time than what we’re facing right now and I believe we can get through this trial in a proper fashion and that’s what I want to do.

Trial resumed on May 8. When the prosecution rested, defense counsel presented testimony from mental-health experts, a childhood friend, a co-worker, and Mc-Manus’s mother and sister, all in an effort to substantiate an insanity defense. The expert witnesses testified that McManus has a low IQ and several mental-health conditions, including depression, attention deficit and hyperactivity disorder (“ADHD”), and a reading learning disability. Court-appointed experts also described McManus’s “low-average” intelligence and mental illnesses, and his IQ tests were entered into the record.

The jury rejected the insanity defense and found McManus guilty. The parties stipulated to incorporate the guilt-phase evidence into the penalty phase of the trial. The defense called one witness, Dr. John Ireland, who offered additional testimony about McManus’s mental illness, learning disability, and low IQ. The jury recommended the death penalty. After weighing the aggravating and mitigating factors, the judge imposed a sentence of death as recommended by the jury.

New lawyers were appointed to perfect McManus’s appeal. They first filed a verified motion to correct errors, arguing that McManus had been incompetent for much of the trial. In support of the motion, the new defense team called two witnesses: Glenn Grampp, one of McManus’s trial attorneys; and Dr. Roger Maickel, professor emeritus of pharmacology and toxicology at Purdue University.

Grampp testified that before the trial began, McManus was capable of understanding the proceedings and participating in his defense, though he had difficulty reading. Things changed dramatically after the panic attack on the first day of testimony. Grampp testified that when he spoke to McManus after he returned from the hospital, “I don’t think he had a clue of what happened earlier in the trial.” From that point onward McManus “provided no assistance whatsover.” Grampp testified that McManus was unresponsive, seemed unaware of what was going on in the courtroom, and for the next two days “just sat slouched over like he was in a stupor.” Grampp saw little improvement in his client’s condition when trial resumed after the one-week recess.

Dr. Maickel testified about the cognitive effects of the medications used to treat McManus during trial. He told the judge that although the drug combination and dosages were nontoxic, their net effect was to turn McManus’s brain into “a neuropsy-chopharmacological soup,” significantly altering his ability to function rationally. Remeron (the antidepressant) and Xanax (the antianxiety drug) each have a sedative effect; Dr. Maickel explained that the effect is more pronounced if the drugs are taken together because each one interferes with the metabolic breakdown of the other. He testified that Xanax is, in fact, classified as a sedative: “the older term used to be minor tranquilizer” and the “prototype drug of that class is Valium.” He explained that Xanax “by itself’ disrupts normal thought processes, producing a general “spaciness” or “fuzziness” and frequent involuntary lapses into “daydreaming.” Dr. Maickel testified that a therapeutic dose of these two drugs in combination would significantly impair the patient’s ability to absorb what’s going on around him and make important decisions: “at best” the patient would be functioning at about 50 percent of normal cognitive capacity. Dr. Maickel also said that it takes at least two to three weeks for a patient to adapt to and become tolerant of these medications.

In a brief bench ruling, the judge denied the motion to correct errors:

[0]n the issue of competency, we were faced with a situation that from the testimony of the physicians and the people involved, that this Defendant was having panic attacks, because he was on trial, in this trial. And the question was if we postpone the trial, and have the trial two weeks or two months from now or two years from now, is there any reason to believe that he would not be having these panic attacks again, because they, apparently, if they were valid, and I have no reason to believe they weren’t, were caused by the trial. What this Court tried to do, then, was to get competent people to either adjust his medication or do whatever it took to get him in good enough shape to be competent to stand trial in this very serious trial, because the alternative would be never to try him, which wasn’t acceptable. And by the time they were done, I was convinced that he was competent.

With that, McManus began his appeals and pursuit of postconviction remedies.

B. Subsequent Procedural History

1. Direct Appeal

The Indiana Supreme Court affirmed the judgment on direct appeal. McManus v. State (“McManus I”), 814 N.E.2d 253 (Ind.2004). On the question of McManus’s competency to stand trial, the court gave the trial judge’s rulings “great deference.” Id. at 260. After reviewing the testimony of Drs. Mohammadi and Whitehead in some detail, the court held as follows:

While the testimony was often equivocal, the consensus of the witnesses was that the medications assisted McManus in participating in his trial. Without the medications, McManus proved unable to cope with the stress of the proceeding. McManus’s situation is markedly different from the defendant who requires medication to attain competence so that the trial can begin. Before trial, Mc-Manus was competent and participated in preparing his case. The administration of medication appeared to manage a sudden onset of stress, rather than to medicate a diagnosed psychosis. Reliance on psychotropic drugs during, trial is obviously to be approached with great care, and competency hearings to evaluate the effects on a defendant’s ability to appropriately participate in his or her defense are very important. In the case at bar, we cannot say that the trial court’s competency determination was clearly erroneous....

Id. at 264.

2. State Postconviction Review

Shortly after McManus was sentenced, the Supreme Court issued its deci sion in Atkins holding that executing the intellectually disabled violates the Eighth Amendment’s prohibition of cruel and unusual punishments. 2 536 U.S. at 321, 122 S.Ct. 2242. Long before Atkins, however, Indiana prohibited the execution of the intellectually disabled. See 1994 Ind. Acts 1851-52 (codified at Ind.Code § 35-36-9-6). Indiana law also establishes a procedure for litigating the question of intellectual disability before trial. See Ind.Code §§ 35-36-9-3, -5. Failure to use the statutory procedure waives the right to raise the claim later. See Smallwood v. State, 773 N.E.2d 259 (Ind.2002).

McManus did not use the statutory procedure for litigating the issue pretrial. Instead, he waited until his postconviction petition to argue that he is intellectually disabled and thus categorically ineligible for the death penalty under Atkins. Shortly before he filed his petition, however, the Indiana Supreme Court held that the state’s statutory procedure did not fully comply with Atkins and modified it accordingly. See Pruitt v. State, 834 N.E.2d 90, 102-03 (Ind.2005) (holding that the statutory requirement that the defendant prove intellectual disability by “clear and convincing evidence” is inconsistent with Atkins and substituting a lower “preponderance of the evidence” burden of proof). The court later held that McManus’s Atkins claim was properly raised by postcon-viction motion because it did not “ripen” until Pruitt modified the statutory procedure. State v. McManus (“McManus II”), 868 N.E.2d 778, 784-85 (Ind.2007).

Atkins largely left to the states the task of developing standards for determining intellectual disability. 536 U.S. at 317, 122 S.Ct. 2242. Indiana uses the following definition: An “ ‘individual with mental retardation’ means an individual who, before becoming twenty-two (22) years of age, manifests: (1) significantly subaverage intellectual functioning; and (2) substantial impairment of adaptive behavior.” Ind. Code § 35-36-9-2. To measure “subaverage intellectual functioning,” the Indiana Supreme Court has adopted the clinical standard used by the American Association on Mental Retardation (“AAMR”) and the American Psychiatric Association (“APA”): a “full-scale IQ test score ... two standard deviations below the mean; i.e., an IQ between 70 and 75.” 3 Woods v. State, 863 N.E.2d 301, 304 (Ind.2007) (cit ing Atkins, 536 U.S. at 309 n. 5, 122 S.Ct. 2242); see also Williams v. State, 793 N.E.2d 1019, 1028 (Ind.2003).

To support his claim, McManus presented testimony from family members, employers, and teachers, but the key witnesses were Dr. Dennis Olvera, a psychologist and expert in intellectual disability, and Dr. Edmond Haskins, a clinical neu-ropsychologist who gave McManus a battery of IQ tests in anticipation of the post-conviction petition.

Dr. Haskins reported that McManus’s test results yielded a full-scale IQ score of 78. Earlier IQ tests from McManus’s childhood and a round of testing administered for purposes of the insanity defense had produced a range of scores from a low of 70 to a high of 81. (We will discuss the IQ scores in more detail in a moment.) At a hearing on the postconviction petition, Dr. Haskins testified that “it’s fair to say [McManus] has an IQ that’s in the high 70s, maybe even low 80s, under the most optimal of conditions.” But he also said it was his “best guess” that McManus was functioning at “a drastically reduced level of adaptiveness and effectiveness” at the time of the crimes because of his other deficits (in particular, his ADHD) and the severe stress he was experiencing due to his impending divorce. Dr. Olvera likewise concluded that although McManus’s IQ scores were too high to meet the clinical standard for intellectual disability, there was a “good possibility” that he was functioning at the level of intellectual disability at the time of the crimes.

The state relied on testimony from Dr. Martin Groff, a psychologist. Dr. Groff did not examine McManus, but he reviewed the relevant record evidence, including the IQ tests, Dr. Haskins’s report, Dr. Olvera’s report, and the reports of the mental-health experts who testified at trial. Dr. Groff testified that McManus did not meet the clinical standard for intellectual disability.

McManus raised several other claims in his postconviction petition; only two are relevant here. McManus reasserted his claim that he was not competent for most of the trial. He also alleged that his trial attorneys were ineffective for failing to present more evidence of intellectual disability during the penalty phase.

The postconviction court found by a preponderance of the evidence that McManus was intellectually disabled and thus ineligible for the death penalty under Atkins. All other claims were rejected. The judge vacated the death sentence and resen-tenced McManus to life without parole.

A divided Indiana Supreme Court reversed the finding of intellectual disability, reinstating the death sentence. See Mc-Manus II, 868 N.E.2d 778. The majority opinion began by reviewing the evidence of McManus’s intellectual functioning. The court noted that of the five IQ tests in the record, three placed McManus’s full-scale IQ above the 70-75 range required to establish significantly subaverage intellectual functioning. Id. at 785-86. Two tests produced full-scale numerical IQ scores above the range: (1) a score of 81 on a test administered when McManus was 11 years old; and (2) a score of 78 on Dr. Haskins’s test, administered at age 34. A third test (at age 7) placed McManus in the “lower limits of [the] low average range,” though no numerical score was reported. Id. at 782, 786. The remaining two tests recorded full-scale IQ scores of 72 (at age 14) and 70 (this test was administered at age 30, while McManus was awaiting trial). Although these scores were within the range for subaverage intellectual functioning, the examiners cautioned that the scores may not accurately reflect McManus’s true IQ because he was not putting forth his maximum effort (during the test administered at age 14) and was anxious and depressed (during the test administered at’ age 30, while he was awaiting trial). Id. at 787.

After recounting this evidence, the state supreme court concluded that the “testing history alone demonstrates McManus is not significantly subaverage as to intellectual functioning.” Id. But the court did not stop its analysis there; the justices also traced the circumstantial evidence bearing on McManus’s intellectual functioning and concluded that it did not support a claim of intellectual disability. The court noted, for example, that McManus graduated from high school, had a positive work history at several jobs, and was able to care for his seriously disabled daughter. Id.

The court then reviewed the evidence of McManus’s adaptive functioning—the second part of the definition of intellectual disability—and concluded that McManus’s scores in conceptual, social, and practical functioning did not reflect substantial impairment in these adaptive-behavior domains. Id. at 788-90. The failure of proof on either component of the definition independently defeated McManus’s claim of intellectual disability. Accordingly, the majority reversed the postconviction court’s finding of intellectual disability, reinstating the death sentence. Id. at 789. Two justices dissented, faulting the majority for not deferring to the findings of the post-conviction court. Id. at 792-93 (Boehm, J., dissenting).

Having addressed the Atkins claim at significant length, the justices swiftly rejected McManus’s remaining claims. As relevant here, the court held that res judi-cata barred McManus from relitigating the issue of his competency to stand trial. Id. at 790. And the court rejected the challenge to trial counsel’s decision not to present additional evidence of McManus’s intellectual impairment during the penalty phase of trial. Id. at 791-92. Substantial evidence on this subject was admitted during the guilt phase of trial (including the IQ evidence) and incorporated by stipulation into the penalty phase; the court held that counsel’s decision not to repeat or bolster this evidence during the penalty phase was not deficient performance. Id.

3. Federal Habeas Review

The case then moved to federal district court. McManus filed a habeas petition under 28 U.S.C. § 2254 raising six claims of constitutional error; four are relevant to this appeal. First, McManus reprised his claim of categorical ineligibility for the death penalty under Atkins. Second, he faulted the state courts for misapplying federal due-process standards regarding his competency to stand trial. He also raised a new claim under Riggins v. Nevada that he was forced to appear before the jury in a “drug-induced stupor” in violation of his right to due process. Finally, he argued that his trial attorneys were constitutionally ineffective because they did not present additional evidence of intellectual disability during the penalty phase of the trial.

The district court denied relief on all claims but granted a certificate of appeala-bility on the Atkins issue. We expanded the certificate to include the. three additional claims we have listed above.

II. Discussion

Federal habeas review of state criminal judgments is highly deferential. AEDPA authorizes federal courts to grant a writ of habeas corpus only when the state-court proceeding “resulted in a decision that was contrary to, or involved an unreasonable application of, clearly established Federal law, as determined by the Supreme Court of the United States” or “resulted in a decision that was based on an unreasonable determination of the facts in light of the evidence presented in the State court proceeding.” § 2254(d)(1)—(2).

A state-court decision is contrary to clearly established federal law when “it applies a rule that contradicts the governing law set forth in [Supreme Court] cases, or if it confronts a set of facts that is materially indistinguishable from a decision of [the Supreme] Court but reaches a different result.” Brown v. Payton, 544 U.S. 133, 141, 125 S.Ct. 1432, 161 L.Ed.2d 334 (2005). An unreasonable application of clearly established federal law occurs when “‘the state court correctly identifies the governing legal principle from [Supreme Court] decisions but unreasonably applies it to the facts of the particular case.’” Emerson v. Shaw, 575 F.3d 680, 684 (7th Cir.2009) (brackets in original) (quoting Bell v. Cone, 535 U.S. 685, 694, 122 S.Ct. 1843, 152 L.Ed.2d 914 (2002)).

“Unreasonable” in this context means more than merely incorrect; a state court’s application of Supreme Court precedent must be “‘so erroneous as to be objectively unreasonable.’ ” Id. (quoting Badelle v. Correll, 452 F.3d 648, 654 (7th Cir.2006)). This standard exceeds even the elear-error standard of review. White v. Woodall, — U.S.-, 134 S.Ct. 1697, 1702, 188 L.Ed.2d 698 (2014). Put differently, a state court’s application of Supreme Court precedent will satisfy reasonableness review if there is room for fair-minded jurists to disagree about it. Quintana v. Chandler, 723 F.3d 849, 855 (7th Cir.2013).

The state court’s factual determinations are cloaked with a presumption of correctness, and the presumption can be overcome only by clear and convincing evidence. See 28 U.S.C. § 2254(e)(1). What this means is that we must be “objectively convinced that the record before the state court does not support the state court’s findings in question.” Ben-Yisrayl v. Davis, 431 F.3d 1043, 1048 (7th Cir.2005); see also Ward v. Sternes, 334 F.3d 696, 704 (7th Cir.2003) (explaining that a state court’s factual determination is unreasonable only if it is “against the clear and convincing weight of the evidence”); Taylor v. Maddox, 366 F.3d 992, 1000 (9th Cir.2004) (explaining that a factual finding is unreasonable under AED-PA if “an appellate panel, applying the normal standards of appellate review, could not reasonably conclude that the finding is supported by the record”). Stated differently, a state court’s decision is factually unreasonable only when it “rests upon fact-finding that ignores the clear and convincing weight of the evidence.” Goudy v. Basinger, 604 F.3d 394, 399 (7th Cir.2010); see also § 2254(e)(1) (requiring clear and convincing evidence to overturn a state-court determination of a factual issue).

We review the district court’s denial of habeas relief de novo. McElvaney v. Pollard, 735 F.3d 528, 531 (7th Cir.2013).

A. Atkins Claim

McManus first challenges the Indiana Supreme Court’s determination that he is not intellectually disabled. He focuses solely on the state high court’s factual findings, apparently conceding that the court reasonably applied the rule of Atkins and that Indiana’s definition of intellectual disability is constitutionally sound.

A claim of factual unreasonableness is difficult to win. To succeed, the petitioner must grapple with the statutory presumption of correctness and the steep burden required to overcome it. And here, the state supreme court comprehensively scrutinized the evidence of intellectual disabili ty before finding it wanting, making Mc-Manus’s burden especially daunting.

As we’ve explained, Atkins largely left to the states the job of developing criteria to determine which death-row prisoners are “so impaired as to fall within the range of mentally retarded offenders” who may not be executed. 536 U.S. at 317, 122 S.Ct. 2242. Even so, the Court noted with approval the accepted clinical definitions of intellectual disability that require both subaverage intellectual functioning and substantial deficits in adaptive skills, both of which must manifest before adulthood. Id. at 318, 122 S.Ct. 2242. More recently the Court held in Hall v. Florida that the general understanding of medical experts will “inform[ ]” but not “dictate” whether a person has an intellectual disability that precludes his execution under the Eighth Amendment. — U.S. -, 134 S.Ct. 1986, 2000, 188 L.Ed.2d 1007 (2014). Hall also mandated that the legal standard for determining subaverage intellectual functioning must account for the margin of error in IQ testing. Id. at 2001.

As we’ve noted, McManus does not attack Indiana’s statutory definition of intellectual disability, which borrows from the criteria used by the medical community and thus is not out of step with either Atkins or Hall. Indiana requires a showing of both “significantly subaverage intellectual functioning” and “substantial impairment of adaptive behavior,” both of which must manifest before the age of 22. § 35-36-9-2. This definition is consistent with the clinical standards promulgated by the APA in the Diagnostic and Statistical Manual of Mental Disorders. See Am. Psychiatric Ass’n, Diagnostic and Statistical Manual of Mental Disorders 33 (5th ed. 2013) [hereinafter DSM-V]. Based on the DSM-V diagnostic criteria, a person is considered to have subaverage intellectual functioning if he scores two standard deviations below the mean on an appropriate intelligence test. Id. at 37. For most IQ tests the mean is 100, the standard deviation is 15, and thus a full-scale IQ score of 70 is the benchmark. See Hall, 134 S.Ct. at 1995-96. Accounting for the standard margin of error, as required by Hall, yields a range not a point: a full-scale IQ score of 70-75 or lower ordinarily will satisfy the first requirement for a finding of intellectual disability. Id. (discussing the standard error of measurement).

The second requirement evaluates impairment of adaptive functioning. The medical community measures adaptive behavior across three domains: conceptual, social, and practical. 4 To satisfy this component of the definition, a person’s adaptive functioning in at least one domain must be “sufficiently impaired that ongoing support is needed in order for the person to perform adequately in one or more life settings at school, at work, at home, or in the community.” DSM-V, supra, at 38. Moreover, the deficits must be caused by the person’s intellectual impairment. Id. The DSM-V requires that the deficits in both intellectual and adaptive functioning appear during childhood or adolescence. Id. at 33, 38.

The American Association on Intellectual and Developmental Disabilities (“AAIDD”) (f/k/a the American Association on Mental Retardation) uses an essentially equivalent definition of intellectual disability: a person must manifest, before the age of 18, “significant limitations in both intellectual functioning and in adaptive behavior as expressed in conceptual, social, and practical adaptive skills.” Am. Ass’n on Intellectual & Developmental Disabilities, Intellectual Disability: Definition, Classification, and Systems of Supports 221 (11th ed. 2010) [hereinafter AAIDD, Intellectual Disability].

Indiana’s definition largely tracks that used by the AAIDD. See Pruitt, 834 N.E.2d at 108. The only difference is the age by which the deficits in intellectual and adaptive functioning must manifest: the statute raises the age from 18 to 22.

After canvassing the record evidence in some detail, the Indiana Supreme Court found that McManus failed to establish that he suffered from either significantly subaverage intellectual functioning or substantial impairment of adaptive behavior. Each of these findings is independently sufficient to defeat his claim of intellectual disability, so if either one holds up under reasonableness review, habeas relief is unwarranted.

1. Intellectual Functioning

The state high court began its analysis by recounting the IQ-seore evidence, noting that McManus was tested on five occasions, at the ages of 7, 11, 14, 30, and 34. McManus II, 868 N.E.2d at 782. The first test placed McManus “-within the lower limits of the low average range,” but the examiner did not record a precise numerical score. Id. (emphasis added). The second test, administered at age 11, recorded a full-scale IQ score of 81, which the examiner also classified as “within a low average range.” Id. McManus scored a 72 on his third IQ test at age 14—within the range of mild intellectual disability—but the court observed that this test was accompanied by a note from the examiner indicating that McManus had not put forth his full effort and the score likely understated his true intellectual ability. Id.

McManus was tested again at age 30, while he was awaiting trial, and achieved a full-scale IQ score of 70. Again, however, the examiner—Dr. Michael Gelbort, a clinical psychologist who was a defense witness at trial—noted that the score likely understated McManus’s true intellectual capacity because he was “anxious and depressed at the time of testing.” Id. The most recent IQ test was administered by Dr. Haskins in February 2006 in connection with the postconviction petition. Mc-Manus was then 34 and achieved a full-scale IQ score of 78. Id.

The state supreme court found it significant that all of the mental-health experts agreed that based on his IQ scores, Mc-Manus was not intellectually disabled. Id. at 786 (“Experts for the trial court, the State, and the defense testified both at trial and during the post-conviction hearing that McManus is not below the level of intellectual functioning that defines mental retardation.”). Dr. Ireland testified that McManus’s true IQ was likely in the “80-type range.” Id. Dr. Gelbort testified that although McManus scored a 70 on the test he administered, he “could probably score slightly higher” if his depression and anxiety abated, though “not significantly so.”

Dr. Haskins agreed that the two lowest scores were likely not indicative of Mc-Manus’s true intellectual functioning. He administered the most recent IQ test, on which McManus achieved a full-scale score of 78. Dr. Haskins explained at the post-conviction hearing that McManus’s true IQ was “in the high 70s, maybe even low 80s, under the most optimal of conditions,” although it was his “best guess” that he was functioning at a lower level at the time of the crimes. Dr. Olvera agreed that Mc-Manus was not intellectually disabled because his IQ score of 78 was too high, but he shared Dr. Haskins’s view that Mc-Manus may have been functioning at the level of an intellectually disabled person at the time of the murders. Finally, Dr. Groff discounted the two lower IQ scores for the reasons noted by the examiners: McManus had not applied his full effort or was anxious and depressed at the time of the test. Id. at 786-87. Dr. Groff concluded that McManus’s IQ scores did not meet the clinical standard for significantly subaverage intellectual functioning.

After cataloging all this evidence, the Indiana Supreme Court found that “Mc-Manus’ testing history alone demonstrates McManus is not significantly subaverage as to intellectual functioning.” Id. at 787. This finding is amply supported by the record. The IQ-test evidence does not place McManus within the range of intellectual disability. Three of the five scores placed him in the “low-normal” or “low-average” range of intellectual functioning, and the two borderline scores were accompanied by examiner notes qualifying the results. On this record, it was objectively reasonable for the state supreme court to discount the two lowest test scores (because they came with qualifiers) and give greater weight to the other IQ scores, as interpreted by the experts. See Thomas v. Allen, 607 F.3d 749, 757 (11th Cir.2010) (holding that the district court’s finding of subaverage intellectual functioning was not clearly erroneous despite an isolated IQ score above the intellectual-disability cutoff).

It’s worth emphasizing that none of the experts testified that McManus falls within the range for a diagnosis of intellectual disability based on the IQ-test evidence as a whole. It’s true that two defense experts—Dr. Haskins and Dr. Olvera—said that McManus may have been functioning at the level of an intellectually disabled person at the time of the crimes based on the combination of his other mental-health deficits and the stress of his looming divorce. 5 But that’s not the relevant question under Atkins. The Supreme Court approved the use of medically accepted clinical criteria defining intellectual disability by reference to impairments in intellectual and adaptive functioning that manifest by the end of the developmental period. 536 U.S. at 308 n. 3, 122 S.Ct. 2242; id. at 317 n. 22, 122 S.Ct. 2242. In other words, intellectual disability for both diagnostic and Eighth Amendment purposes is not transitory; it’s a chronic condition based on symptoms that manifest before adulthood. To accept the testimony of Drs. Haskins and Olvera as a basis to find McManus ineligible for the death penalty would require an extension of Atkins, not an application of it.

McManus argues that the older IQ scores may have overstated his performance because of the so-called “Flynn Effect,” which refers to the increase in IQ scores over time. IQ tests are scored on a scale that is relative to the population. Test developers determine the mean and standard deviation relative to the population at the time the test is developed; this is referred to as “norming” the test. Be cause IQ scores rise over time, the Flynn Effect posits that the mean score will rise above 100 until the test is re-normed. See Geraldine W. Young, Note, A More Intelligent and Just Atkins: Adjusting for the Flynn Effect in Capital Determinations of Mental Retardation or Intellectual Disability, 65 Vand. L.Rev. 615, 617 (2012). For example, a person with a measured IQ of 76 on a test normed in 1990 would not be two standard deviations below the mean if he took the test the year it was normed. But if the same person took the same test in 2010, a score of 76 might amount to a measure of intellectual functioning two standard deviations below the mean because the average IQ score would have risen by about six points (0.3 points per year multiplied by 20 years). 6

The Flynn Effect is taking on increased prominence in habeas litigation alleging death ineligibility under Atkins. See Frank M. Gresham & Daniel J. Reschly, Standard of Practice and Flynn Effect Testimony in Death Penalty Cases, 49 Intellectual & Developmental Disabilities 131 (2011). The circuits are not consistent in their approach on this point. Compare, e.g., Black v. Bell, 664 F.3d 81, 95 (6th Cir.2011) (faulting state court for not considering the Flynn Effect under Tennessee law) and Walker v. True, 399 F.3d 315, 322-23 (4th Cir.2005) (finding the Flynn Effect relevant to whether someone is two standard deviations below the mean), with Hooks v. Workman, 689 F.3d 1148, 1170 (10th Cir.2012) (“Atkins does not mandate an adjustment for the Flynn Effect.”). See also Thomas, 607 F.3d at 757-58 (collecting cases and noting that no expert consensus exists on how to apply the Flynn Effect to individual cases); Young, Adjusting for the Flynn Effect, supra, at 631-41 (analyzing the different approaches used in state and federal courts); Gresham & Reschly, supra, at 136-37 (criticizing those administering psychological tests for failing to consider the Flynn Effect). Our circuit has not yet weighed in.

Although the Flynn Effect is acknowledged in the field, it is not common practice to adjust IQ scores by a specific amount to account for the phenomenon. Hooks, 689 F.3d at 1170. More to the point here, nothing in Atkins suggests that IQ test scores must be adjusted to account for the Flynn Effect in order to be considered reliable evidence of intellectual functioning. The Indiana Supreme Court found as a factual matter that McManus does not suffer from significantly subaverage intellectual functioning based on the IQ evidence; that finding is well supported by the record, including the testimony of the expert witnesses, all of whom agreed that his scores do not place him within the clinical range for intellectual disability. The court’s failure to consider the Flynn Effect does not make its factual determination objectively unreasonable.

McManus also argues that the court’s decision was based on stereotypes and other misconceptions about people with intellectual disabilities. More specifically, he takes issue with two parts of the court’s analysis. First, he claims that the court wrongly attributed his two lowest test scores to his learning disability and ADHD. Second, he argues that the court should not have placed any weight on the evidence that he graduated from high school, successfully worked three jobs, and took care of his profoundly disabled child. He notes in particular that although he graduated from high school, he was in a special-education curriculum, and his jobs were in unskilled labor. Finally, he maintains that no evidence supports the proposition that intellectually disabled parents cannot care for disabled children.

We do not doubt that intellectually disabled people graduate from high school (with' or without the assistance of special-education programming) and also hold down jobs. And we accept that some intellectually disabled parents have the capacity to care for a disabled child. But McManus overstates the Indiana Supreme Court’s use of this circumstantial evidence. The court did not draw any firm conclusions about McManus’s intellectual functioning from this evidence alone. Rather, the court mentioned this evidence only in passing and only as additional support for its conclusion that McManus’s higher IQ scores reflected his true intellectual ability. See McManus II, 868 N.E.2d at 787.

Similarly, the state supreme court did not conclude that McManus’s lower test scores were exclusively attributable to his learning disability and ADHD. Under current diagnostic criteria, intellectual disability and specific learning disorders may explain low testing performance. See DSM-V, swpra, at 67 (“The learning difficulties are not better accounted for by intellectual disabilities.... ”). IQ tests have “performance” and “verbal” components, and the test-taker’s scores on each component combine to produce a full-scale IQ score. With only one exception, Mc-Manus consistently achieved a performance score in the 80s, but his verbal scores were lower. 7 This suggests that Mc-Manus’s full-scale IQ scores were pulled down by his poor verbal performance. It’s reasonable to infer that McManus’s learning disability and ADHD contributed to his lower verbal scores, and indeed Drs. Ireland and Haskins, testified to that effect.

In short, the Indiana Supreme Court’s factual determination that McManus’s intellectual functioning is not significantly subaverage is solidly grounded in the record and thus is not objectively unreasonable. This holding alone is independently sufficient to reject McManus’s Atkins claim. For completeness, however, we move to the state supreme court’s assessment of McManus’s adaptive functioning.

2. Adaptive Behavior

As noted, Indiana has adopted the AAIDD’s criteria for measuring substantial impairment of adaptive behavior, which requires “significant limitations ... in conceptual, social, and practical adaptive skills.” 8 AAIDD, Intellectual Disability, supra, at 221; McManus II, 868 N.E.2d at 788. To quantify and measure McManus’s adaptive skills, Dr. Olvera administered the Adaptive Behavior Assessment System II (“ABAS-II”). One of his assistants administered a second test, the Vineland-II Adaptive Behavior Scales (“VABS-II”).

The ABAS-II arrives at composite scores in the domains of conceptual, social, and practical skills by evaluating ten sub-domains. The conceptual domain has three subdomains: (1) communication; (2) functional academics; and (3) self-direction. The social domain has two subdo-mains: (4) leisure and (5) social. And the practical domain has five subdomains: (6) community use; (7) home/school living; (8) self-care; (9) health and safety; and (10) work. McManus II, 868 N.E.2d at 788 n. 8. Dr. Olvera scored the ABAS-II based on interviews with people who knew Mc-Manus well: his mother and three of his employers. Id. at 788. Each domain has a mean score of 100 and a standard deviation of 15, so (like IQ tests) a score at or below 70 will signal intellectual disability. For scores in the subdomains, the mean is 10 and the standard deviation is 3; accordingly, a person is considered to be in the range of intellectual disability if he scores 4 or lower, two standard deviations below the subdomain mean. Id. at 788 n. 8.

McManus achieved composite scores of 82 in the conceptual domain, 90 in the social domain, 93 in the practical domain, for a General Adaptive Composite score of 88 on the ABAS-II. None of these scores falls within the intellectually disabled range. Id. at 788. McManus’s subdomain scores, which formed the basis for the composite scores, likewise do not fall within the range commonly associated with intellectual disability. Eight of the subdo-main scores ranged from 8-12, with an average of 9.5, well within the average range. A ninth score was lower—a 5 in community use—but still above the threshold for intellectual disability. McManus’s only subdomain score below the threshold was a 2 in functional academics.

In contrast, McManus’s composite score on the VABS-II fell well within the range of an intellectually disabled person. The VABS-II test is a bit different; it evaluates the domains of communication, daily living, and socialization, and each category has three subdomains. 9 Id. at 789. The composite scores have a mean of 100 and a standard deviation of 15, so again, a score at or below 70 will, signal intellectual disability. Like the ABAS-II, the VABS-II test is based on interviews with people familiar with the subject, though for this test Dr. Olvera’s assistant interviewed only McManus’s sister and sister-in-law. Mc-Manus achieved a composite score of 75 in the domain of daily living, 71 in socialization, and 21 in communication skills, for an overall composite score of 55.

Faced with contradictory testing data, the Indiana Supreme Court credited the ABAS-II results. The court found that this testing instrument “most closely resembles the AAMR definition” of intellectual disability, which the Indiana statute largely mirrors. Id. at 788. The court also noted that the VABS-II data was obtained through interviews with Mc-Manus’s family members only, so the results might have been skewed by “the affection of the relatives who supplied the input.” Id. at 789. Finally, the court noted that McManus’s dismal communication score of 21 on the VABS-II had a substantial impact on his overall composite score on that test. Dr. Olvera testified that a communication score that low would indicate the presence of a severe intellectual disability; no .one had suggested that McManus suffers from an intellectual disability of that magnitude. This anomaly, the court held, was an additional reason to be suspicious of the VABS-II results. Id.

McManus argues that the court erred in disregarding the VABS-II test scores. We see two problems with this argument. First, nothing in Atkins commands the use of a particular test or clinical instrument for determining whether a person is intellectually disabled. See 536 U.S. at 316,122 S.Ct. 2242. Indiana’s high court had the discretion to find the ABASII a more discerning measure of adaptive behavior. Second, McManus’s argument ignores the deferential standard of review. We have no authority to second-guess the state supreme court’s resolution of a conflict in the testing evidence; we are authorized to grant habeas relief only if the state court’s determination of the facts is against the clear weight of the evidence. That’s hardly the ease here. The Indiana Supreme Court made a considered judgment about which test results were more reliable and gave sound reasons, grounded in the evidence, for crediting one test over the other.

Finally, McManus zeroes in on his low scores in the subdomains of functional academics and community use. Dr. Groff characterized the functional-academics score as “clearly” within the range of intellectual disability and the community-use score as “close.” But a low score in a single subdomain of adaptive behavior is not clear and convincing evidence of intellectual disability under any commonly accepted definition. Moreover, with his diagnosis of a reading disorder, it’s not surprising that McManus would have a low score in functional academics while scoring within (or near) the low-average range in other areas of adaptive behavior.

In sum, the Indiana Supreme Court made an objectively reasonable factual determination that McManus is not intellectually disabled and thus not categorically ineligible for the death penalty under Atkins.

B. Competency

McManus also argues that the state courts unreasonably applied federal due-process principles in addressing his competency to stand trial. We agree.

“[T]he Constitution does not permit trial of an individual who lacks mental competency.” Indiana v. Edwards, 554 U.S. 164, 170, 128 S.Ct. 2379, 171 L.Ed.2d 345 (2008) (internal quotation marks omitted). A person is competent to stand trial when “he has sufficient present ability to consult with his lawyer with a reasonable degree of rational understanding [ ] and ... a rational as well as factual understanding of the proceedings against him.” Dusky, 362 U.S. at 402, 80 S.Ct. 788; see also Edwards, 554 U.S. at 170, 128 S.Ct. 2379; Drope v. Missouri, 420 U.S. 162, 171-72, 95 S.Ct. 896, 43 L.Ed.2d 103 (1975); Benefiel v. Davis, 357 F.3d 655, 659 (7th Cir.2004).

The due-process rule announced in Dusky has deep common-law origins and implements the fundamental principle that it is unjust to punish a person who lacks the mental capacity to understand the proceedings against him and participate in his own defense. Drope, 420 U.S. at 171, 95 S.Ct. 896 (“It has long been accepted that a person whose mental condition is such that he lacks the capacity to understand the nature and object of the proceedings against him, to consult with counsel, and to assist in preparing his defense may not be subjected to a trial.”); see also 4 William Blackstone, Commentaries *24-25. The trial court must conduct a competency hearing—sua sponte if necessary—when there is substantial reason to doubt the defendant’s mental fitness to stand trial. Pate, 383 U.S. at 385, 86 S.Ct. 836; Sturgeon v. Chandler, 552 F.3d 604, 612 (7th Cir.2009). Whether a competency hearing is warranted is necessarily an individualized determination. “Relevant factors include any evidence of irrational behavior, the defendant’s demeanor in court, and any medical opinions on the defendant’s competency to stand trial.” Sturgeon, 552 F.3d at 612.

Again, because this is a habeas case, we do not apply these standards directly. Rather, we ask whether the state court’s decision was contrary to, or an unreasonable application of, clearly established federal law, or was based on an unreasonable determination of the facts. See § 2254(d)(l)-(2).

It’s undisputed that McManus’s panic attacks and the medications used to treat them raised bona fide doubts about his competency. The panic attacks were severe enough to require two trips to the hospital, and every medical doctor who testified in this case agreed that the drugs used to treat McManus have significant cognitive effects, most notably on perception, judgment, and (in the case of Versed) memory. Dr. Maickel, the pharmacologist, testified that the medications in combination turned McManus’s brain into “a neuropharma-cological soup.” We do not need to go that far to accept that the medications raised substantial doubt about McManus’s mental fitness to proceed. Indeed, everyone agrees that a competency inquiry was necessary. The disputed question is whether the state courts reasonably applied the federal due-process framework for adjudicating competency questions. We conclude that they did not.

We begin with the trial judge’s rulings. The defense team repeatedly moved for a mistrial or a continuance, arguing that Mc-Manus’s panic attacks and the medications prescribed to control them rendered him incompetent to proceed. The judge summarily denied each motion. The first of these rulings contains faint echoes of the Dusky standard, so we can safely assume that the judge was aware of the constitutional mínimums. But the judge’s rulings do not reflect a process of reasoning tied to the legal standard; they are entirely conclusory. And the record does not supply the missing premises, largely because the judge never ordered a formal competency examination.

Recall that after the second panic attack and hospital visit, the judge summoned Dr. Mohammadi, the emergency-room physician, to testify about McManus’s condition. Dr. Mohammadi told the court that Mc-Manus was calmer and more rational than he was in the emergency room. But he also said that McManus’s mental processes were significantly slowed by medication. Importantly, Dr. Mohammadi did not give an expert opinion about McManus’s competency to stand trial. He wasn’t asked to—he was an emergency-room physician, after all, not a forensic psychiatrist. And indeed, he qualified his testimony at every turn, acknowledging the limits on both the scope of his expertise and his examination of McManus; he agreed with defense counsel that McManus would need to be seen by a psychiatrist to determine an appropriate medication regimen that would allow him to understand and participate in the trial. Dr. Mohammadi’s testimony was at once equivocal and contingent and does not provide an evidentiary foundation for us to conclude that the judge’s summary decision implicitly addressed the federal competency standard.

The judge’s ruling on the verified motion for a mistrial was even more problematic. This time the judge did have testimony from a specialist: Dr. Whitehead, a psychiatrist. But Dr. Whitehead was not asked to—and in fact did not—conduct a forensic competency examination. To use the judge’s own words, Dr. Whitehead was brought in to get McManus “fixed up” enough to complete the trial, or as Dr. Whitehead put it more diplomatically, he was there to “help McManus feel better.” And Dr. Whitehead qualified his testimony in two crucial respects: (1) although Mc-Manus was pleasant and cooperative, he spoke and moved slowly, so the doctor had difficulty obtaining useful information from him; and (2) McManus had a panic attack in the middle • of the interview, so Dr. Whitehead could not complete the examination.

Again, the judge summarily denied the defense motion, never, addressing whether McManus was competent to proceed or even hinting at the elements of the applicable legal standard. Instead, the judge simply asserted his belief that Mc-Manus’s condition was either self-induced or the result of stress from the trial and would not improve if a mistrial or continuance were granted. Whether McManus’s condition would “improve” is not the right question; it does not address whether Mc-Manus was presently competent to proceed in the sense required to comply with minimum 'standards of due process. The Dusky standard is not relative to a person’s normal functioning; it is a minimum objective threshold that must be exceeded.

In the end, we cannot escape the fact that despite substantial evidence of Mc-Manus’s decompensation and the powerful cognitive effect of the psychotropic medication he was taking, the judge failed to directly address the competency standard and never made a clean factual finding that McManus had a rational and factual understanding of the proceedings and a present rational capacity to consult with his lawyers about his defense. And because no competency examination was ordered (and thus no expert testified to the elements of the legal standard), we cannot infer that the judge’s ruling was implicitly keyed to the appropriate legal test. The judge might have conducted his own on-the-record colloquy with McManus to check his understanding of the proceedings and his capacity to assist in his defense. If that had happened, we might be able to conclude that the judge made a reasonable independent judgment about McManus’s competency. After all, a defendant’s competency to stand trial is a legal inquiry, not a medical inquiry, and “the judge is the expert on what mental capabilities the litigant needs in order to be able to assist in the conduct of the litigation.” Holmes v. Buss, 506 F.3d 576, 581 (7th Cir.2007). But that did not happen here. 10

The judge’s failure to order a proper competency evaluation is itself problematic. Because there were bona fide doubts about McManus’s mental fitness, the Indiana Code required the court to appoint two or three medical experts having expertise in determining competency and hold an appropriate hearing. See Ind. Code § 35-36-3-1. 11 Indiana’s procedures are facially sufficient to satisfy due process, see Drope, 420 U.S. at 173, 95 S.Ct. 896 (approving an analogous Missouri statute), but it’s clear the judge wasn’t using the statutory procedure. One of the doctors who testified did not have the required expertise, and neither doctor conducted an appropriate examination. Dr. Mohammadi was an emergency-room physician whose expertise was in stabilizing patients. Dr. Whitehead was a treating psychiatrist whose qualifications to assess competency were not placed in the record. We can probably assume that he had the relevant expertise, but he was not asked to evaluate McManus for competency and did not in fact do so. Instead, Dr. Whitehead was brought in to get McManus “fixed up” to finish the trial.

Although the judge failed to follow Indiana’s statutory procedure, that’s not a basis for federal habeas relief. See Wilson v. Corcoran, 562 U.S. 1, 131 S.Ct. 13, 14, 178 L.Ed.2d 276 (2010) (“Federal courts may not issue writs of habeas corpus to state prisoners whose confinement does not violate federal law.”); Drope, 420 U.S. at 172, 95 S.Ct. 896 (“The Court did not hold [in Pate ] that the procedure prescribed by. [Illinois law] was constitutionally mandated, although central to its discussion was the conclusion that the statutory procedure, if followed, was constitutionally adequate.” (citations omitted)). A competency hearing may be constitutionally adequate yet fall short of Indiana’s statutory requirements. For the reasons we have explained, however, the judge did not conduct a constitutionally adequate competency hearing.

The trial court’s failure to adjudicate the competency question under the standards established in Dusky, Pate, and Drope becomes significant for our review of the Indiana Supreme Court’s decision. As we have noted, the state high court deferred to the trial judge’s rulings. McManus I, 814 N.E.2d at 260, 264. By subjecting a constitutionally inadequate trial-court decision to deferential review, the Indiana Supreme Court did not adequately vindicate the federal due-process interests at stake. See Harrison v. McBride, 428 F.3d 652, 666-67 (7th Cir.2005).

Moreover, the state supreme court’s analysis was itself incomplete. Although the court recited the correct due-process standard early in its decision, McManus I, 814 N.E.2d at 260-61, the court never actually applied it. After acknowledging that the doctors’ testimony was equivocal, the court held that the “consensus of the witnesses was that the medications assisted McManus in participating in his trial.” Id. at 264. Reasonable minds can differ about whether the record fairly supports that interpretation. But asking whether the medications were “assisting” McManus does not resolve the competency question, at least not without further factual and legal analysis. The due-process inquiry asks whether the defendant had a present factual and rational understanding of the trial proceedings and the capacity to assist his lawyers with a reasonable degree of rational understanding. See Dusky, 362 U.S. at 402, 80 S.Ct. 788. Because the court never actually applied this standard, it too committed unreasonable error.

Accordingly, although the standard of review under AEDPA is deferential, the record does not permit a conclusion that the state courts reasonably applied federal due-process standards in adjudicating Mc-Manus’s competency to stand trial. 12 Mc-Manus prevails on this claim.

This brings us to the question of remedy. Sometimes a retrospective competency hearing is possible, though for obvious reasons contemporaneous determinations are preferred. Young v. Walls, 311 F.3d 846, 848 (7th Cir.2002). A retrospective hearing may be a remedial option if “it is still possible to hold a meaningful retrospective hearing to determine if the defendant was fit to stand trial at the time of the original state proceedings.” Estock v. Lane, 842 F.2d 184, 188 (7th Cir.1988). “The passage of even a considerable amount of time may not be an insurmountable obstacle if there is sufficient evidence in the record derived from knowledge contemporaneous to trial.” United States ex rel. Bilyew v. Franzen, 686 F.2d 1238, 1247 (7th Cir.1982) (quoting United States v. Makris, 535 F.2d 899, 904 (5th Cir.1976)). Compare Burt v. Uchtman, 422 F.3d 557, 566 (7th Cir.2005) (granting the writ of habeas corpus where medication disrupted a defendant’s alertness during trial), with Young, 311 F.3d at 848-49 (noting that “when a defendant’s condition is stable, evidence adduced after trial allows a reliable reconstruction of the defendant’s mental state at trial”).

Here, the problem extends far beyond the passage of time. McManus’s condition was- highly unstable; the panic attacks might have resulted from changes in his medication, difficulties that would not be observable after he was stabilized. And regardless of what caused the attacks, the drugs administered to curtail them clearly affected McManus’s cognition during trial. We cannot see how n'ew testimony before the district court could possibly provide the necessary information to retrospectively assess his competency under the applicable legal standard. Habeas relief is warranted. 13

III. Conclusion

For the foregoing reasons, McManus is not entitled to habeas relief on his claim of intellectual disability under Atkins. But the state courts unreasonably applied federal due-process standards in adjudicating McManus’s competency to stand trial. Accordingly, we REVERSE the district court’s judgment and REMAND with instructions to grant the writ unless Indiana gives notice of its intent to retry McManus within a reasonable time to be fixed by the district court.

1 . From the transcript:

Q: So for the next four and a half hours, there’s no reason that he wouldn't—if somebody got up here and said I saw him commit the crime, he would know what they were saying?

A: Yes.

Q: And he would know whether it was true or not?

A: I would believe so, yes.

2 . Atkins and earlier opinions used the term "mental retardation,” not "intellectual disability” or "intellectual developmental disorder,” the preferred terms used today. See Hall v. Florida,-U.S.-, 134 S.Ct. 1986, 1990, 188 L.Ed.2d 1007 (2014); Am. PsychiatI ric Ass’N, Diagnostic and Statistical Manual of Mental Disorders 33 (5th ed.2013) [hereinafter DSM-V]. We will follow the Supreme Court’s lead in Hall and use the term "intellectual disability” rather than "mental retardation,” although some references to the older term cannot be avoided because that is the term used in previous court decisions and the relevant Indiana statutes, Ind.Code §§ 35-36-9-1, et seq.

3 . A score of 70 is two standard deviations below the mean IQ test score of 100. Hall, 134 S.Ct. at 1994-95. The five-point range accounts for the standard margin of testing error. See id. at 1995-96; see also Atkins v. Virginia, 536 U.S. 304, 309 n. 5, 122 S.Ct. 2242, 153 L.Ed.2d 335 (2002) (”[A]n IQ between 70 and 75 or lower ... is typically considered the cutoff IQ score for the intellectual function prong of the mental retardation definition.”). In Hall the Supreme Court struck down a Florida rule requiring an IQ score of 70 or below for a claim of intellectual disability; the Court held that a hard cutoff of 70 is too rigid because it fails to account for the standard margin of error in IQ testing. See 134 S.Ct. at 2001. Indiana law accounted for a 5-point margin of error long before Hall made this a constitutional requirement. See State v. McManus ("McManus II"), 868 N.E.2d 778, 785-86 (Ind.2007); Woods v. State, 863 N.E.2d 301, 304 (Ind.2007); Williams v. State, 793 N.E.2d 1019, 1028 (Ind.2003).

4 . The conceptual domain "involves competence in memory, language, reading, writing, math reasoning, acquisition of practical knowledge, problem solving, and judgment in novel situations, among others.” DSM-V, supra note 2, at 37. The social domain examines interpersonal skills, such as communication, empathy, and social judgment. Id. And the practical domain inquires into a person’s ability to manage his life, such as money management, behavior, and job responsibilities. Id. The DSM-IV requires deficits in at least two of the following areas: communication, self-care, home living, social/interpersonal skills, use of community resources, self-direction, functional academic skills, work, leisure, health, and safety. Am. Psychiatric Ass'n, Diagnostic and Statistical Manual of Mental Disorders 41 (4th ed.2000).

5 . They couched their testimony in nonscientific terms as a “best guess” (Dr. Haskins) and a “good possibility” (Dr. Olvera) that McManus was ftmctioning at a lower intellectual level at the time of the crimes.

6 . James Flynn, the eponym of the "Flynn Effect” theory, estimated that IQ scores increase at 0.3 points per year. James R. Flynn, The Mean IQ of Americans: Massive Gains 1932 to 1978, 95 Psychol. Bull. 29, 32-34 (1984).

7 . His three other performance scores were 81, 85, and 88. See McManus II, 868 N.E.2d at 786.

8 . The DSM-V, like the AAIDD, now looks to the conceptual, social, and practical domains. The older list of skill areas has been subsumed into these categories. See DSM-V, supra note 2, at 37.

9 . These subdomains are: receptive, expressive, and written (for communication); personal, domestic, and community (for daily living); and interpersonal/relations, play and leisure, and coping (for socialization).

10 . On. May 9 defense counsel engaged in a short colloquy with McManus to make a record about his decision not to testify. A second brief colloquy occurred the next day, during the sentencing phase of trial, on the same subject. The judge relied on these colloquies as support for his decision to deny the motion to correct errors. This is problematic for several reasons. The colloquies were brief; they were conducted by counsel, not the court; and they were not addressed to the competency standard. Moreover, given the powerful effects of the medication, the colloquies on May 9 and 10 did not answer whether McManus was competent during the earlier phases of the trial.

11 . Section 35-36-3-1 of the Indiana Code provides:

If at any time before the final submission of any criminal case to the court or the jury trying the case, the court has reasonable grounds for believing that the defendant lacks the ability to understand the proceedings and assist in the preparation of a defense, the court shall immediately fix a time for a hearing to determine whether the defendant has that ability. The court shall appoint two (2) or three (3) competent, disinterested:

(1) psychiatrists;

(2) psychologists endorsed by the Indiana state board of examiners in psychology as health service providers in psychology; or

(3) physicians; who have expertise in determining competency.

12 . A slightly different way of looking at it is that a legal error infected the trial court’s fact-finding process, so the resulting factual determination is unreasonable. See Taylor v. Maddox, 366 F.3d 992, 1001 (9th Cir.2004) (“Obviously, where the state court’s legal error infects the fact-finding process, the resulting factual determination will be unreasonable and no presumption of correctness can attach to it.”). In other words, the state supreme court affirmed an unreasonable factual finding on deferential review.

13 . Our holding on the competency issue makes it unnecessary for us to address Mc-Manus's remaining claims, which allege other constitutional errors at trial.