Lawrence Jefferson v. GDCP Warden

941 F.3d 452

Opinion

[PUBLISH]

IN THE UNITED STATES COURT OF APPEALS

FOR THE ELEVENTH CIRCUIT

________________________

No. 17-12160

Non-Argument Calendar

________________________

D.C. Docket No. 1:96-cv-00989-CC LAWRENCE JOSEPH JEFFERSON, Petitioner - Appellee,

versus GDCP WARDEN, Respondent - Appellant.

________________________

Appeal from the United States District Court for the Northern District of Georgia

________________________

(October 17, 2019)

Before ED CARNES, Chief Judge, and TJOFLAT and MARCUS, Circuit Judges. MARCUS, Circuit Judge:

Lawrence Jefferson was convicted of felony murder in 1986 and sentenced to death by a Georgia jury. Since then, his challenge to his death sentence has included state collateral proceedings, one previous trip in this Court, an appeal to

2

the Supreme Court, and two proceedings in federal district court. Following this considerable procedural history, we find ourselves in an uncommon situation, resolving a petition for habeas corpus filed before the enactment of the Antiterrorism and Effective Death Penalty Act of 1996, Pub. L. No. 104-132, 110 Stat. 1214 (1996) (the “AEDPA”), and applying pre-AEDPA law to an issue expressly formulated for us by the Supreme Court. The crux of Jefferson’s claim is one in which we are well-versed -- that he received ineffective assistance of counsel during the sentencing phase of his trial because his lawyers failed to adequately investigate his mental health, and, in particular, whether he suffered from organic brain damage at the time of the killing. Jefferson argues that he was prejudiced by counsel’s error because it is reasonably probable that at least one of the jurors would not have sentenced him to death if the evidence of his severe mental impairment had been presented. Georgia law requires a life sentence unless the jury unanimously agrees that the defendant should die.

Before we may address this claim, however, we are obligated to resolve whether the district court properly concluded under pre-AEDPA law that the state habeas court deprived Jefferson of a full and fair hearing when it denied his petition, thus stripping the state court’s factual determinations of a presumption of correctness -- the question the Supreme Court has directed us to answer. After thorough review, and extensive fact-finding by the district court, we conclude that

3

the state habeas court’s fact-finding was not entitled to deference in the pre- AEDPA regime. The state habeas court adopted verbatim the State’s proposed order; it offered no guidance to the Assistant Attorney General drafting the proposed order, including how to resolve important credibility conflicts; apparently, it did not review the order, other than signing it, dating it, and changing the concluding sentence, notwithstanding the glaring errors it contained; and it did so ex parte without so much as affording Jefferson a chance to challenge any of it or propose an alternative order.

Having determined that the state habeas court’s findings are not entitled to a presumption of correctness, we come to the more common habeas inquiry -- whether, on the factual record compiled by the district court, Jefferson suffered ineffective assistance of counsel at the sentencing phase of his trial. We think the district court correctly determined that Jefferson’s trial lawyers’ conduct fell beneath an objective standard of reasonableness when they ignored the unambiguous written recommendation of their retained psychologist that a neuropsychological evaluation be conducted in order to rule out an organic etiology and explain Jefferson’s mental health and behavior at the time he committed the homicide. They also ignored a series of red flags that suggested that Jefferson’s aberrant behavior was the result of organic brain damage sustained at the age of two when his head was run over by an automobile.

4

Finally, in light of the substantial evidence Jefferson put forward showing that he suffers from organic brain damage that significantly affected his conduct and impulse control at the time of the killing, we conclude that the district court did not err in finding that Jefferson has been prejudiced by his lawyers’ deficient performance. Among other things, the jury heard nothing about the extent of the head injury Jefferson sustained when he was struck in the head by an automobile as a two-year-old child; nothing about his five-day hospitalization or the headaches and blackout spells he thereafter suffered; and nothing about the resulting frontal lobe and neurological damage he sustained so early in his life, which likely caused diminished impulse control, irritability and short-temperedness, intermittent outbursts of rage, impaired judgment, and an inability to foresee the consequences of his actions. All of this is to say that the jury was presented with a profoundly misleading picture of Jefferson’s sentencing profile and moral culpability because the most important mitigating circumstances were withheld. Indeed, the most powerful explanation for an otherwise inexplicable crime -- that Jefferson suffered from organic brain damage that severely impaired his judgment and his ability to control his behavior -- was never presented to the jury. Thus, we affirm the judgment of the district court and grant Jefferson’s habeas petition. He is entitled to a new sentencing proceeding.

I.

5

A.

In March, 1986, Lawrence Jefferson was found guilty of felony murder that occurred during the commission of an armed robbery. Jefferson v. Hall, 570 F.3d 1283, 1290 (11th Cir. 2009), cert. granted, judgment vacated sub nom. Jefferson v. Upton, 560 U.S. 284 (2010), and vacated and remanded sub nom. Jefferson v. Warden, Ga. Diagnostic & Classification Prison, No. 07-12502, 2010 WL 3431652 (11th Cir. July 21, 2010). The basic facts surrounding the murder and Jefferson’s trial and state habeas proceedings, which have already been detailed many times, are not contested, and we summarize them only briefly today.

According to testimony elicited at Jefferson’s trial, the body of the murder victim, Edward Taulbee, was spotted lying in the woods near Lake Allatoona in North Georgia on May 2, 1985 by passing motorists. Id. at 1287. An autopsy revealed that Taulbee sustained two lacerations above his left eyebrow, one laceration above his forehead, one laceration above his right ear, five lacerations on the back of his scalp, several fractured teeth, an abrasion across his face, an abrasion across his back, skull fractures, brain bruises, and brain hemorrhaging. Id. The medical examiner concluded that these head injuries had caused Taulbee’s death and that he had died sometime between 5 and 11 p.m. on May 1. Id.

The ensuing police investigation discovered that Taulbee worked at Zenith Construction Company, as did Jefferson. Id. at 1287-88. In a police interview,

6

Jefferson said he had seen Taulbee leave their construction worksite around 5:30 p.m. on May 1 to go fishing. Id. at 1288. Jefferson also claimed he left shortly thereafter, went home, and spent the evening at his apartment. Id.

The police learned from other Zenith employees that Taulbee was known to carry large sums of money that he lent to coworkers when they needed cash. Id. When police subsequently interviewed Jefferson’s neighbors, they discovered that, contrary to Jefferson’s report, Jefferson had gone with Taulbee in Taulbee’s car after work on May 1 to fish at Lake Allatoona. Id. One neighbor added that Jefferson had returned from this outing alone and on foot, and, the next day, he gave her an ATM card to hide. Id. Another said that he went to Jefferson’s apartment on the evening of May 1, and Jefferson was acting “funny” and his chest looked red. Id. at 1289. Jefferson also mentioned that his “little fat buddy was dead,” an obvious reference to Taulbee. Id. A third neighbor revealed that he had driven Jefferson, upon request, to Lake Allatoona on the night of May 1, and that once there Jefferson briefly went into the woods and returned with a fishing pole, which he threw back into the woods. He also gave a tackle box to the neighbor for safekeeping. Later that night, Jefferson tried to make a withdrawal at an ATM machine. Id. at 1289-90.

Thereafter, police found Taulbee’s car in a parking lot near Jefferson’s apartment, and learned that the ATM card and tackle box Jefferson had given to his

7

neighbors had belonged to Taulbee and that someone had tried to use Taulbee’s ATM card in the early morning hours of May 2. Id. at 1289-90. Jefferson was arrested on May 7, 1985. When police asked him why he had killed Taulbee, oddly he replied that he didn’t need to be around other people, that he wanted to be executed, and that he wanted to be put to sleep. Id. at 1290.

Jefferson was indicted in August, 1985 in the Superior Court of Cobb County, Georgia, and convicted by a jury in March, 1986. At the penalty phase of Jefferson’s trial, the State introduced evidence of Jefferson’s past offenses, including a series of armed robberies he had committed in Louisville, Kentucky. Id. In mitigation, Jefferson offered the testimony of two Cobb County police officers, who testified that Jefferson had not caused problems while he was incarcerated. Id. Jefferson’s mother testified that Jefferson had had a hard upbringing because he grew up without a father and had to help take care of his siblings, and that he was responsible, generous, gentle, kind, and playful as a child. Id. at 1291. She also mentioned that Jefferson’s head had been run over by a car when he was two years old, but she did not elaborate on this incident and was not questioned at all about what effect the injury had on his behavior. Id. Jefferson’s sister corroborated that Jefferson was generous and responsible, and had helped care for his siblings when he was younger. Id. Finally, the mother of Jefferson’s

8

children testified that Jefferson was a good father, was responsible and helped care for their children. Id.

Jefferson also testified. He discussed his difficult upbringing without a father, how he helped his mother raise his siblings, and how he left high school to make money for his family. Id. As for the evidence of prior crimes he was said to have committed, he generally denied having been involved in any wrongdoing. Id.

Jefferson’s trial counsel, Marc Cella and Stephen Schuster, presented closing arguments. Their arguments developed themes that Jefferson, while a deeply flawed individual, also had some good in him and was capable of self- improvement; that, because of his difficult upbringing in a poor black family, Jefferson was not as responsible for his actions as most individuals; and that Jefferson deserved mercy. In addition to these arguments, counsel also suggested, albeit only briefly, that Jefferson had not actually killed Taulbee and that lingering doubt about his guilt should sway the jury’s imposition of punishment.

Following deliberations, the jury unanimously recommended that Jefferson die, and the state court sentenced him to death. The jury found two aggravating factors: the offense of murder was committed while the defendant was engaged in the commission of another capital felony (i.e., armed robbery), Ga. Code Ann. § 17-10-30(b)(2) (1986); and that the offense of murder “was outrageously or wantonly vile, horrible, or inhuman” in that it involved an aggravated battery to the

9

victim, id. § 17-10-30(b)(7). Jefferson appealed and the Georgia Supreme Court affirmed his conviction and sentence. Jefferson v. State, 353 S.E.2d 468 (Ga. 1987).

B.

Jefferson sought collateral review in Georgia state court. His petition included a claim that he had received ineffective assistance of trial counsel because his lawyers had failed to reasonably investigate the serious head injury he had sustained as a two-year-old child.

During an evidentiary hearing, the state court took evidence concerning the investigative efforts by Jefferson’s trial counsel, Cella and Schuster, into his background. The court heard that prior to trial, defense counsel had contacted Petitioner’s family in Louisville, Kentucky, and traveled to Kentucky to interview family members. Counsel also retained clinical psychologist Dr. Gary E. Dudley to conduct a mental health examination of Jefferson. Dr. Dudley submitted a report opining that Jefferson was in the midrange level of intelligence and that any mental deficiencies were “quite mild and in no way impair[ed] his judgment or his decision-making capacity.” Dr. Dudley recounted, however, that during one day of testing, Jefferson refused to cooperate with Dudley, berating him for not making an appointment with him, but several days later, Jefferson had apologized to Dudley and explained, “I get that way sometimes, man.” Dr. Dudley’s report noted that

10

“[o]ne possibility that could not be explored because of [Jefferson’s] incarceration has to do with the sequelae to head injury experienced during childhood [and] [i]n my opinion it would be worthwhile to conduct neuropsychological evaluation of this individual to rule out an organic etiology.”

However, attorney Schuster testified that when he later conferred with Dr.

Dudley about his written report, Dr. Dudley had “basically” said that “[t]here is nothing I could testify to that would help you, if you get to the sentencing phase.” Schuster recalled that Dudley worked in Florida in prison psychology and labeled Jefferson as “just a criminal.” As for Dudley’s recommendation of neuropsychological testing, Schuster recounted that Dudley “basically said it may be a waste of time because the rest of the report . . . said that he was non psychotic, that he was intelligent, he had a fairly high I.Q., and . . . ‘[b]ecause he denies the alleged crime.’” According to Schuster, “Dr. Dudley didn’t pursue it and we saw no reason to further pursue it.”

Jefferson introduced mental health evidence at the state habeas proceeding that had not been developed at his sentencing. He offered an affidavit from Dr. James Merikangas, a neurologist who gave him a series of tests and concluded that it was substantially likely “that any violent behavior [Jefferson] may have engaged in was a result of the . . . brain damage” he had suffered as a child. Jefferson also presented an affidavit from Dr. David Price, a clinical psychologist who performed

11

neuropsychological testing on him. He opined that “Jefferson has brain damage, his brain cannot operate in the way a non-damaged brain operates, and this brain damage in and of itself causes abnormal behavior in [] Jefferson over which he has no or substantially limited control.”

Jefferson also submitted an affidavit from Dr. Dudley contradicting Schuster’s recollection of their interaction. In it, Dr. Dudley denied telling Jefferson’s counsel that a neuropsychological evaluation would not be necessary or that it would not be worthwhile. Dudley claimed that he understood that the testing was not requested because funds were not available for it. Dr. Dudley denied telling Schuster that Jefferson was “just a criminal,” or that he could not help Jefferson’s case at the sentencing phase.

Jefferson also introduced affidavits from his mother and his brother. His mother, Vera Jefferson, described the head injury he had sustained at age two, when “a car hit him while he was playing in the alley [and] drug [sic] him across the alley.” She said that Jefferson had “stayed in the hospital for a long time before they let him go home. He ran heavy fevers that whole time, and the people at the hospital told me he was very bad hurt.” After that, his mother “saw changes in [Jefferson],” and “[h]e’d seek out quiet little places to hide a lot.” “As he got up in age a little bit,” she noted that “he was slower than the others to pick up on things,” and “around [the age of nine or ten], [Jefferson] started getting the

12

headaches he’s been having ever since. I used to worry about him driving because he got dizzy and his sight got blurry sometimes.” She said that before the trial, Jefferson’s “attorneys didn’t ask me questions about what I’ve said above though, and if they had I would have told them.”

Jefferson’s brother provided similar testimony, observing that Jefferson “wasn’t a normal type kid[,] [h]e was a nervous type person[,] [h]e got emotional real quick, and he cried real easy,” and that “[h]e used to sweat all the time[,] [e]ven in the cold winter time.” His brother added that Jefferson “got headaches bad,” was “always the sickest one in our family,” and “got dizzy a lot and he couldn’t keep his balance good.” His brother offered that he had not been “contacted by any of Lawrence’s attorneys before his trial.”

After holding evidentiary hearings, the state habeas court issued an opinion that, as it turns out, was a verbatim adoption of an order proposed by the State of Georgia. The opinion set forth factual findings, including that: (1) Jefferson’s trial counsel, Cella and Schuster, had conducted an extensive pretrial investigation into Jefferson’s background, knew that Jefferson had suffered a head injury at a young age, and never saw any evidence that Jefferson had brain damage; (2) after Jefferson’s lawyers retained Dr. Dudley to examine Jefferson, they did not seek any further neuropsychological evaluation, because when Schuster spoke with Dudley, “Dr. Dudley led Mr. Schuster to believe that further investigation would

13

simply be a waste of time because Petitioner was not psychotic, was intelligent and had a fairly high I.Q.” and “Dudley had described Petitioner to Mr. Schuster as ‘just a criminal’ and indicated he had no helpful testimony”; and (3) in light of Dudley’s report and the follow up conversation, “counsel saw no need to have further neuropsychological testing done, particularly given Petitioner’s ‘adamant’ assertion that he did not commit the crimes.”

Following its findings of fact, the state habeas court concluded that “counsel made a strategic decision, based upon their investigation, not to present any type of mental defense or pursue additional neuropsychological testing as counsel had investigated this line of inquiry but found nothing helpful to the defense.” The opinion added that the court had “specifically credit[ed] the testimony of Mr. Schuster and Mr. Cella with regard to their efforts to investigate Petitioner’s mental condition, both as to potential defenses to the crimes themselves as well as mitigating evidence, and their decision not to call Dr. Dudley as a witness in mitigation.” The opinion ultimately found no ineffective assistance of counsel: “This Court has found counsel made a reasonable investigation into the mental health area, obtained an independent report which was not favorable to the defense, followed up on the suggestion of further neurological testing but then were led to believe it would not be worthwhile, had no other evidence indicating further

14

examination was warranted and simply concluded that this line of investigation was no longer plausible.”

The Georgia Supreme Court, reviewing the trial court’s habeas order, also rejected his claim of ineffective assistance of counsel. Jefferson v. Zant, 431 S.E.2d 110, 112-14 (Ga. 1993). Georgia’s high court noted that counsel had “prepared extensively for a possible sentencing phase, including travelling to Jefferson’s home in Kentucky to interview family members and obtaining a psychological evaluation by a psychologist recommended to them . . . .” Id. at 113. The court highlighted Dudley’s report, which had opined that Jefferson “was impulsive, immature, had poor impulse control and would ‘pursue immediate gratification without properly considering possible consequences,’” but “was not psychotic or seriously mentally ill.” Id. The court then said:

The report did suggest that a neuropsychological evaluation might be “worthwhile . . . to rule out an organic etiology.” However, in subsequent telephone conversations with Jefferson’s attorneys, the psychologist indicated that further examination would not likely be beneficial because Jefferson did not have serious mental problems and was of normal intelligence. Counsel explained that in light of this advice, as well as their own observations of Jefferson and Jefferson’s adamant assertion that he had not committed the crime, they concluded that there was no reason to explore further any mental issue, or to call the psychologist to testify in mitigation. . . .

Jefferson’s counsel investigated the possibility of mental mitigation and concluded that there was nothing substantive to develop. The record supports the finding that Jefferson’s counsel did not believe and had no reason to believe, that Jefferson suffered mental deficiencies that could have been exploited in mitigation.

15

Id.

C.

Jefferson filed his Petition for Writ of Habeas Corpus in the United States District Court for the Northern District of Georgia on April 23, 1996, and filed an amended petition on August 27, 1998. The first time Jefferson’s case was before the federal habeas court, the court granted in part and denied in part Petitioner’s “corrected amended petition for a writ of habeas corpus.” The court granted Petitioner relief as to “Petitioner’s claim that trial counsel was ineffective during the penalty phase of trial,” while denying all of the other claims raised in the petition. The district court determined that Jefferson’s trial counsel had performed ineffectively when they failed to conduct a reasonable investigation of mitigating evidence, including, in particular, mental health evidence.

On appeal, this Court reversed. Jefferson v. Hall, 570 F.3d 1283 (11th Cir.

2009), cert. granted, judgment vacated sub nom. Jefferson v. Upton, 560 U.S. 284 (2010), and vacated and remanded sub nom. Jefferson v. Warden, Ga. Diagnostic & Classification Prison, No. 07-12502, 2010 WL 3431652 (11th Cir. July 21, 2010). We reasoned that Jefferson’s lawyers had conducted an adequate investigation, and, in particular, were justified in declining to pursue further neuropsychological testing or other inquiries into Jefferson’s mental health. In reaching this conclusion, we relied in large measure on the state court’s

16

unambiguous findings that Dr. Dudley had made oral statements to the effect that further testing would likely be a waste of time and that there was nothing he could offer that would help Jefferson at sentencing. See id. at 1309. In addition, we noted that “Jefferson’s counsel had other substantial reasons for deciding not to pursue neuropsychological testing” that were supported by the state court’s factual findings, including that “neither Schuster nor Cella witnessed anything in their frequent and extensive dealings with Jefferson to indicate that he was afflicted with any mental disorder that might have been relevant to the murder”; that “Jefferson had unwaveringly proclaimed his innocence”; and that “the State’s case against him suffered from many weaknesses.” Id. at 1304-05. Throughout, our decision relied on the factual findings made by the state habeas court, which we said were entitled to a “a strong presumption of correctness.” Id. at 1306.

Because we’d held that Jefferson’s lawyers’ performance was objectively reasonable, we did not consider whether Jefferson was prejudiced by his lawyers’ failure to conduct additional investigation into his mental health. Id. at 1309. After agreeing with the state habeas courts’ rejection of Jefferson’s ineffective- assistance-of counsel claim, we reversed the district court’s order granting in part Jefferson’s petition and remanded for the district court to reinstate Jefferson’s original sentence. Id. at 1311.

17

Judge Carnes, however, dissented. Id. at 1311 (Ed Carnes, J., dissenting).

He reasoned that, even accepting the factual determinations made by the state habeas court as correct, trial counsel’s failure to have further testing conducted was unreasonable: “Knowing that his client had suffered a serious head injury as a child and that a psychologist had issued a formal report recommending further testing to determine if the injury had caused organic brain damage, no reasonable attorney would have let the matter drop because of informal remarks made to him during a telephone conversation.” Id. at 1313. Judge Carnes also explained that Jefferson was prejudiced by his lawyers’ deficient performance, since “the jury was deprived of the singularly most powerful explanation for the crime -- Jefferson suffered from organic brain damage that impaired his judgment and his ability to control his behavior.” Id.

Thereafter, the Supreme Court granted certiorari and vacated this Court’s opinion and remanded the case. Jefferson v. Upton, 560 U.S. 284 (2010). The Supreme Court did not address the merits of the ineffective-assistance-of-counsel matter (that is, whether defense counsel’s conduct fell beneath an objective standard of reasonableness, nor whether the defendant was prejudiced by counsel’s claimed deficiency), but instead simply asked whether this Court and the district court should have afforded deference to the state court’s findings of fact under the prior version of 28 U.S.C. § 2254(d) that applies to Jefferson’s case. Under the

18

pre-AEDPA statute, a state habeas court’s factual determinations are not entitled to a presumption of correctness under certain circumstances, including if they are not “fairly supported by the record”; if “the factfinding procedure employed by the State court was not adequate to afford a full and fair hearing”; if “the applicant did not receive a full, fair, and adequate hearing in the State court proceeding”; or if “the applicant was otherwise denied due process of law in the State court proceeding.” Id. at 291 (citing 28 U.S.C. § 2254(d)(2), (6), (7), (8) (1994)). The Supreme Court questioned whether the state habeas court’s verbatim adoption of the prosecution’s proposed order implicated the statute and stripped the state court’s factual findings of deference.

In laying out its concerns, the Supreme Court noted that, while the verbatim adoption by a court of findings of fact prepared by a prevailing party is often permissible, the use of such a practice might be procedurally problematic “where (1) a judge solicits the proposed findings ex parte, (2) does not provide the opposing party an opportunity to criticize the findings or to submit his own, or (3) adopts findings that contain internal evidence suggesting that the judge may not have read them.” Id. at 294. The Supreme Court declined to decide “in the first instance” whether, in light of these problems, “any of the exceptions enumerated in §§ 2254(d)(1)-(8) apply in this case, especially given that the facts surrounding the state habeas court’s process are undeveloped.” Id. (citation omitted). Instead

19

the Supreme Court remanded the case to the lower courts in order to assess “the precise nature of what transpired during the state-court proceedings” in order to determine “whether the state court’s factual findings warrant a presumption of correctness, and to conduct any further proceedings as may be appropriate in light of their resolution of that issue.” Id. We, in turn, remanded the case to the district court to perform the necessary fact-finding.

D.

On remand, the district court gathered evidence to address: (1) whether the state habeas court afforded Jefferson a full, fair, and adequate hearing consistent with due process, 28 U.S.C. § 2254(d) (1994); and (2) whether, based on the new evidence submitted at the federal hearing, Jefferson received ineffective assistance of counsel in violation of the Sixth Amendment at the sentencing phase of his trial. Jefferson v. Sellers, 250 F. Supp. 3d 1340 (N.D. Ga. 2017).

1.

Several people were deposed concerning the adequacy of the state court procedures: (1) Paula Smith, Assistant Attorney General and lead counsel for Respondent during the state habeas proceedings; (2) J. Christopher Desmond, who had been an attorney with Schreeder, Wheeler, and Flint and served as Jefferson’s pro bono counsel during the state habeas proceedings; (3) Elizabeth Vila Rogan, an attorney with the Georgia Resource Center who provided support to Desmond; and

20

(4) Wendell Boyd English, a law clerk in the Waycross Judicial Circuit who assisted the state habeas corpus judge, Joseph B. Newton, who passed away in 2000.

These essential facts were adduced before the district court on round two.

Jefferson’s case was heard in the Superior Court of Butts County by Judge Newton, who had never handled a habeas corpus action involving a capital case previously. While the case was before him, Judge Newton held a two-day evidentiary hearing during which the parties presented exhibits, affidavits, and live testimony concerning, among other things, Jefferson’s claim of ineffective assistance of counsel at the penalty phase of trial. At the conclusion of the hearing, the state habeas court set out a briefing schedule for post-hearing briefs, giving the Petitioner 90 days from the completion of the evidentiary hearing transcript to file his brief; and giving the State 30 days thereafter to respond. Nothing was mentioned about a reply brief; the judge simply said he’d “like to have everything in by the end of September.” Petitioner filed his post-hearing brief on October 23, 1991, and Respondent filed a post-hearing brief on December 24, 1991. Jefferson, 250 F. Supp. 3d at 1345.

Sometime before August 20, 1992, Wendell Boyd English, in his capacity as a law clerk assisting Judge Newton, and at the judge’s direction, called Assistant Attorney General Smith, and asked the State to prepare a proposed final order

21

denying Petitioner state habeas relief. English worked for all three judges in the Waycross Judicial Circuit; as a law clerk for the circuit, he primarily worked for Judge Elie Holton, but assisted Judge Newton with two habeas cases, one of which was Jefferson’s. Smith did not recall if English gave her further instructions about what to include in the order, and English testified that Judge Newton did not give him any instructions about how Smith was to draft the order. Smith did not consider the request unusual; she recalled that it was “standard procedure” at that time for judges to ask the prevailing party to draft an order in state habeas cases. But, Rogan, who had served on Jefferson’s habeas team, testified she had never, before or after, had any experience with a judge simply requesting a proposed order from the respondent without also requesting a proposed order from the petitioner.

On August 20, 1992, Smith mailed English the proposed final order she’d prepared with a cover letter. Jefferson, 250 F. Supp. 3d at 1347. Smith also sent copies of the letter and order to Jefferson’s counsel, Desmond.1 When asked whether she’d had the capacity to e-mail the order to Judge Newton or send him a disk, Smith responded that she sent him a hard copy. There is no evidence that

1 We commend Assistant Attorney General Smith for bringing into the open the ex parte nature of the communication by sending copies of the letter and order to Jefferson’s counsel, thereby affording opposing counsel the opportunity to object and to challenge the order on that basis. Without having done so, the matter might never have been disclosed to counsel or the Court. Assistant Attorney General Smith acted in the best tradition of the Bar.

22

Smith had any conversations or other communications with Judge Newton or anyone else on his behalf about the contents of the proposed final order.

While English did research and read the transcript of the hearing and any other documents forwarded to Judge Newton, he did not remember seeing any post-hearing briefs, or having any lengthy discussion with Judge Newton about the case. Nor did English recall ever receiving or reviewing the proposed order or doing any research on it. When he testified that he did not remember seeing the proposed order (despite a letter in the file indicating that it had been sent to him), English said he thought it would have been sent directly to Judge Newton, and admitted, “I’m flabbergasted by this. . . I would not have done that today.” He did not explain what he would “not have done.” He added that he did not recall Judge Newton ever asking him to review the order or to do any drafting with respect to the order. He had nothing more to do with matter once Assistant Attorney General Smith submitted it.

Petitioner’s state habeas counsel, Desmond, first learned of Judge Newton’s request that Respondent’s counsel prepare a proposed final order on August 24, 1992, when he received Smith’s letter with the enclosed proposed final order for Judge Newton’s signature. Neither Desmond nor Rogan previously had been aware of any type of ex parte communication between the state habeas court and

23

Assistant Attorney General Smith after the evidentiary hearing was held on April 22 and 23, 1991.

Desmond responded by filing a “Motion to Recuse Hearing Judge and a Motion to Reveal Ex Parte Communications” on August 28, 1992. Counsel objected that Judge Newton had asked only one side to submit proposed findings of fact and conclusions of law. Desmond added that if he had been allowed to do so, Jefferson would have offered competing findings of fact and conclusions of law. Desmond explained that he “wanted to memorialize in the record the evident bias and lack of care and interest in this judge doing his job. And I specifically used the phrase, you know, ‘as is apparent from the course of events to date, however, submission of such a proposed order from me would be an exercise in futility as it’s evident that the Court has already made up its mind of the outcome of this case.’”

Judge Newton summarily denied the recusal motion on September 28, 1992, without ever inviting Petitioner’s counsel to file a competing proposed order or competing proposed findings of fact and conclusions of law. Two days after denying the recusal motion, Judge Newton executed an order that, except for the concluding sentence, the date, and his signature on the final page, was identical to the proposed order drafted by the State. During the subsequent appeal of that order

24

to the Georgia Supreme Court, the parties stipulated “that the state habeas corpus judge adopted the proposed order as his own.”

Judge Newton’s order did not correct many misspellings, grammatical errors, or any miscitations to cases. Jefferson has since identified twenty-one errors in the order, including a misspelling on the very first page:

1. Page 1, line 12: aggravated “battery” misspelled as aggravated “batty”;

2. Page 5, line 9: “constitutionally” misspelled as “constitutonally”;

3. Page 9, lines 13-17: the text indicates that the indictment was returned on August 8, 1985 and that Schuster was appointed “one day after the indictment was returned,” which would be August 9, 1985, but the order incorrectly lists the date as May 9, 1985;

4. Page 11, line 19: “state’s” should be “state”;

5. Page 19, line 5: “was” should be “as”;

6. Page 19, line 4: “was” should be “were” to correspond to the plural noun “defenses”;

7. Page 20, line 2: “Petitioner” should be “Petitioner’s”;

8. Page 21, lines 6-7: “the any” should be either “the” or “any,” but not both;

9. Page 21, lines 8-9: incorrect usage of “state’s”;

10. Page 21, line 19: “reserved” misspelled as “reservd”;

11. Page 22, lines 15-16: the case Cook v. State, 255 Ga. 565 has an incorrect Southeastern citation, listed as 340 S.E.2d 891, when the actual citation should be 340 S.E.2d 843;

25

12. Page 22, line 21: the correct pincite to Strickland v. Washington, 466 U.S. 668, is pages 688-689 and not 668, the first page of the decision;

13. Page 23, line 3: the word “prospective” in the Strickland quote should be “perspective”;

14. Page 23, line 13: “statements” is misspelled as “statemets”;

15. Page 26, line 2: “institute” should be “constitute”;

16. Page 29, lines 23-25: presumably the court would have found “no attorney error” not simply “no attorney”;

17. Page 30, line 23: “counsel’s” misspelled as “counse’s”;

18. Page 34, line 16: “crimes” misspelled as “crimed”;

19. Page 35, line 12: “remarks” misspelled as “remakrs”

20. Page 36, lines 4-5: the case Conner v. State, 251 Ga. 113, is identified as having been decided in the year “198” but was decided in the year “1983”; and

21. Page 37, lines 25-26: the case Daugherty v. Dugger, 839 F.2d 1426 is incorrectly cited as Daugherty v. Dugger, 839 F.2d 1246.

In addition, both the proposed order and the final order referred to the affidavit of an attorney, Michael Hauptman, who was never contacted in connection with this case, and no affidavit from him was ever submitted to the court. 2 2.

2 According to the state court’s order, Hauptman’s affidavit said that trial counsel’s strategy was unreasonable, but the order rejected this suggestion since “Hauptman could not possibly know all the non-record information which was discovered by [Schuster] and [Cella] and relied upon by them in making their decisions in this case.”

26

The federal district court also gathered its own facts about the mental health investigation that had taken place before and after Jefferson’s trial. It held two evidentiary hearings at which Jefferson’s trial counsel, Cella and Schuster, testified, as did four mental health experts: Dr. Gary Dudley, Dr. James Merikangas, Dr. David Price, and Dr. Steven Macciocchi. Jefferson, 250 F. Supp. 3d at 1354-55. Merikangas and Price were retained by Jefferson’s state habeas counsel, and Dudley was originally retained by Jefferson’s trial counsel. The State presented Dr. Macciocchi to rebut Jefferson’s expert witnesses. The district court’s hearings focused primarily on the dispute concerning whether Petitioner’s attorneys were told orally by Dr. Dudley that an investigation of Jefferson’s traumatic head injury was unnecessary and would be a waste of time, and the conflicting expert evidence presented by both sides concerning the state of Jefferson’s mental health. Id. at 1355.3 Cella was the first witness. Defense counsel testified that he had been aware of Jefferson’s head injury as he was preparing for trial because Jefferson’s mother had told him about it. However, Cella did not follow up with her about the injury.

3 Beyond resolving these disputed facts, the district court relied on the undisputed factual determinations made by the state habeas court in analyzing Jefferson’s ineffective assistance of counsel claim. Id. at 1355 n.7 (“While the Court’s focus in this Order is on the mental health investigation, the Court remains aware of the entirety of trial counsel’s efforts in investigating mitigating evidence in preparation for the penalty phase of Petitioner’s trial. In evaluating the performance of trial counsel, the Court considers the entire scope of the background investigation, which is fully set forth in this Court’s prior Order and in the Eleventh Circuit’s Opinion and altered only slightly by the facts this Court has found upon remand.”).

27

Cella recalled that Jefferson was evaluated by Dr. Dudley before trial, but at the time, Cella’s impression was that Dudley’s evaluation did not turn up any information that would have been especially useful for Jefferson at trial. Cella did not remember following up on Dudley’s written recommendation that Jefferson should be tested for an organic etiology, nor did he remember speaking with Dudley about whether any further evaluation was needed, nor, finally, was he aware of Schuster ever having spoken with Dudley about Dudley’s evaluation. Cella said his failure to follow up on Dudley’s report was due to “inexperience,” and he explained that “I should have followed up on that. And had I had more -- if this had not been my first death penalty case, I expect I would have.”

The district court also took testimony from defense counsel Stephen Schuster. Schuster had practiced law for about ten years before representing Jefferson, and had helped prosecute a few death penalty cases, but Jefferson’s was his first capital case as lead defense counsel. He testified that he had seen no indication of mental health problems in his interactions with Jefferson. And when Schuster and Cella spoke to Jefferson’s family about his upbringing, they decided it had been “relatively normal.”

Schuster retained Dr. Dudley to conduct a psychological evaluation of Jefferson, and his impression, after orally discussing the report with Dudley, was that no further evaluations would be worth conducting. Schuster claimed that

28

Dudley told him that Jefferson was “a sociopath or psychopath and not to worry about it,” and that “Mr. Jefferson was the way Mr. Jefferson was.” Schuster maintained that “we would have asked to do [further testing] if we had been told by Dr. Dudley, you need to do this, seriously need to do this, as opposed to a throw-out in the report.” Schuster also noted that residual-doubt was part of his strategy at sentencing. On cross-examination, Schuster acknowledged that he learned before trial that his client had been hit in the head by a car as a young child.

The Petitioner also called Dr. Merikangas to offer expert mental health testimony. Merikangas studied medicine at John Hopkins University, and did specialty training in neurology and psychiatry at the Yale University School of Medicine. He also served as a chief resident in neurology at Yale. He was board certified in Psychiatry in 1974 and board certified in Neurology in 1978. He has some 25 years of experience specializing in neurology and psychiatry, and served on the Yale faculty and on the American Board of Psychiatry and Neurology. Many of Merikangas’s academic publications have focused on brain injury and brain disease, he was named a Distinguished Life Fellow of the American Psychiatric Association, as well as a Fellow and Founding Director of the American Neuropsychiatric Association, and he served as an examiner for the American Board of Psychiatry and Neurology.

29

Merikangas conducted a comprehensive neurological examination of Jefferson in May, 1989. As part of the exam, he measured Jefferson’s head, which was above average in circumference, suggesting that Jefferson “may have had some brain swelling.” The tests, which were not “specifically designed to demonstrate, evaluate, and quantify brain damage,” also indicated that Jefferson had high blood pressure, asymmetrical deep tendon reflexes, and decreased sensitivity to pinpricks.

Dr. Merikangas reviewed Jefferson’s medical records and education records too, as well as affidavits from Jefferson’s family and the medical reports of Drs. Price and Dudley. According to Merikangas, Dudley’s findings, like the discrepancy between Jefferson’s verbal and performance intelligence (“IQ”) test scores suggested “that [there] may be some brain damage.” Merikangas agreed with Dudley’s recommendation that further neuropsychological testing was needed. Dr. Merikangas added that Dr. Price found right hemisphere and frontal lobe brain damage. Merikangas emphasized that all of the information and testing he relied on would have been available in 1985 and 1986 when Jefferson was indicted and tried.

Based on all the relevant materials, Dr. Merikangas opined that he had “no doubt” that Jefferson suffers from brain damage, “probably localized in the right hemisphere in the frontal lobes of the brain.” Merikangas testified that the likely

30

cause of the brain injury was the car accident. According to Merikangas, a brain injury like Jefferson’s can make someone short-tempered and irritable; prone to headaches; unable to exercise sound judgment or control his impulses; and unable to foresee the consequences of his actions. Merikangas observed that people with brain injuries have the “potential for transient outbursts of rage that are totally inconsistent with” their ordinary behavior, and that symptoms can worsen with time.

Merikangas opined in his affidavit in the state habeas court that the neurological damage “has had a profound and lifelong impact on Mr. Jefferson’s behavior.” Dr. Merikangas said that “[t]he damage produce[d] features including attention deficit disorder, learning disabilities, affective instability, markedly diminished impulse control, transient outbursts of rage which are totally inconsistent with his normal behavioral pattern, impaired social judgment, significant cognitive impairment, apathy, indifference, suspicion, and paranoid ideation. Mr. Jefferson’s brain damage is clearly of the acquired, post-traumatic type. Persons with such severe cognitive disabilities have reduced moral culpability for acts generally, are more likely to fall victim to substance abuse, and, when intoxicated, are certain to be much more symptomatic.”

In federal district court, Merikangas reiterated that Jefferson’s injury made him less able to control his behavior than healthy individuals, and thus, less

31

morally culpable for his criminal misconduct. Jefferson also had “a diminished capacity to conform his conduct to the requirements of the law.” As Merikangas explained, if “[Jefferson] hit someone over the head with a log, it might have been a momentary fit of rage, which was not as controllable by someone who is impaired as it would be in a normal person, who would have simply spoken to that man and had worked it out in other ways.” Merikangas offered that the absence of psychopathic or sociopathic tendencies in Jefferson made it all the more likely that any occasional, violent behavior was the result of a closed head injury and brain damage.

On cross-examination, Merikangas admitted that his forensic involvement in death penalty cases was almost exclusively on the defense side, that he has only testified as an expert for the prosecution on at most two occasions, that he has personal objections to the death penalty, and that he has given many speeches about how to defend against a death sentence. The government also elicited that Merikangas’s evaluation was based on relatively limited information, and that he had not reevaluated Jefferson since the initial evaluation in 1989.

The petitioner also offered the testimony of Dr. Gary Dudley. He received a Ph.D. in clinical psychology from the University of Miami and practiced as a psychologist for forty years. Dudley was contacted in 1985 by Jefferson’s lawyers and retained to perform a psychological evaluation. The examination consisted of

32

a general intelligence test and a couple of personality tests, and revealed that Jefferson scored 94 on the verbal IQ test and 108 on the performance portion of the test. Dudley agreed with Merikangas that the discrepancy between these scores was significant, and opined that it “was instrumental in [his] recommendation to pursue neuropsychological assessment” because a discrepancy like this suggests brain impairment. Dudley noted that Jefferson had a learning disability, which could be caused by a brain injury, and that Jefferson was likely to be impulsive and not prone to giving forethought to his actions, both of which are symptoms of frontal lobe damage. Dudley said that during one of his visits with Jefferson the defendant became noticeably agitated and confused for no apparent reason. He had provided the results of his tests to Jefferson’s lawyers in a written report.

Dudley testified that he had little recollection of the conversations he had with Cella and Schuster at the time of Jefferson’s trial. However, he confirmed that he had “suggested that neuropsychological evaluation would be appropriate” to determine if Jefferson suffered from brain damage of some kind. He flatly denied ever telling anyone on Jefferson’s defense team that neuropsychological testing would be a “waste of time”; he never “back[ed] off” of his written recommendation that further testing should be conducted; he never “retreat[ed]” from that recommendation; or ever reported that the testing was not necessary. Nor did he recall ever telling Jefferson’s lawyers that he would not be of any help

33

to them at sentencing. Rather, Dudley explained that he “assumed that additional testing would be done and then that testimony would be relevant rather than mine.” He also denied ever telling Jefferson’s lawyers that Jefferson was “just a criminal.” Finally, Dudley testified that the testing later conducted by Dr. David Price is just what he had in mind when he recommended further neuropsychological testing.

On cross-examination, Dudley acknowledged that he had no recollection of any specific conversation he may have had with Jefferson’s lawyers about the contents of his report. But on redirect he insisted that he would not have included any recommendation in his written report that he didn’t stand by and that he wouldn’t have told Jefferson’s lawyers that further testing would be a waste of time.

The petitioner also offered the expert testimony of Dr. David Price. Price received a Ph.D. in clinical psychology from Auburn University and received specialized training in treating brain injuries. With extensive experience as an expert witness, he has testified for both sides in criminal cases. Price conducted neuropsychological testing on Jefferson in 1991, including administering the Wechsler Adult Intelligence Scale, the Wide Range Achievement Test, the Denman Neuropsychology Memory Scale, the Halstead-Reitan Category Test, and the Luria-Nebraska Neuropsychological Battery, all of which were considered state

34

of the art neuropsychological tests at the time Jefferson took them. 4 A few of these tests indicated that Jefferson suffered from organic dysfunction. Notably, Jefferson scored at a fourth-grade level in spelling and a sixth-grade level in math on the intelligence tests, suggesting that he had not benefited from his 11th grade education as much as would have been expected. Another test indicated that Jefferson has some memory impairment.

In addition to the testing, Price reviewed affidavits from Jefferson’s family, school and hospital records, and the submissions from Dudley and Merikangas. Price noted that Jefferson’s family members reported that Jefferson would cry as a child because of severe, chronic headaches and had blackout spells where he would actually become unconscious, and which may have been petit mal seizures. Jefferson’s school records showed that he had a low IQ. Price was troubled by the hospital records of Jefferson’s hospitalization following his automobile accident, finding that a five-day hospital stay was a long time and was further indication that Jefferson had sustained a severe traumatic head injury.

4 Dr. Price’s testing regime differed from the testing that Dr. Merikangas conducted. The district court explained that “[a]s a neuropsychologist, Dr. Price quantified Petitioner’s brain functions to give a catalog of performance, whereas Dr. Merikangas, as a neurologist, was equipped to use hands-on, physical testing to ascertain causes.” Jefferson, 250 F. Supp. 3d at 1360. Merikangas articulated the difference this way: “The psychological testing [done by Price], a lot of it is interview or talking to the person and pencil and paper kinds of tests . . . most of it . . . is thought testing rather than the hands-on physical testing, which [a] neurologist does. [A] neurologist, being a medical doctor, is equipped to look for lots of medical and traumatic conditions, whereas the neuropsychologists are simply quantifying what’s there without necessarily in regard to what the causes are or to treatment.”

35

Like Dr. Merikangas, Price found the IQ discrepancy in Dudley’s testing to be highly significant. According to Price, when one hemisphere of the brain is operating more efficiently than the other, there may be an organic problem or brain dysfunction.

Upon completing his tests and reviewing the relevant materials, Price opined that Jefferson had suffered a brain injury and had chronic organic dysfunction. Jefferson suffered from frontal lobe damage that likely was caused by the automobile accident when Jefferson was a two-year-old child. Price had opined in his affidavit in state habeas court that Jefferson’s “brain damage has significant and far reaching implications for [him] because such damage compromises [a] patient’s ability to act in a normal and acceptable manner, often without warning or explanation.” Indeed, said Price, the sequelae from Jefferson’s head jury would have “profoundly alter[ed] Mr. Jefferson’s ability to plan and coordinate his actions, to be aware of the consequences of his behavior, and to engage in premeditated or intentional acts.” In district court, Price again explained that an organic brain injury like Jefferson’s could cause someone to have problems with self-control inhibition, irritability, judgment, planning and understanding consequences. In the case of Jefferson’s brain injury, Price believed that Jefferson suffered from a substantial loss of self-control, which typically results in agitated behavior.

36

By way of example, Price explained that Jefferson had refused to cooperate with Dudley on one occasion, which is an indication of what happens to the Petitioner in novel and stressful situations and is fully consonant with a substantial loss of self-control.

Price also explained that a brain injury suffered by a two-year-old is demonstrably more significant than one in an adult because a two-year-old’s brain is continuing to develop. A two-year-old’s brain injury has substantially more potential to alter cognitive development and cause impairment in neurometabolic function, communication, and development of the frontal lobes.

Finally, Dr. Steven Macciocchi, the State’s rebuttal expert witness, testified.

Macciocchi has a Ph.D. in clinical psychology from the University of Louisville and completed a one-year internship and a one-year fellowship specializing in clinical neuropsychology at the University of Virginia School of Medicine. He has worked as an assistant professor in neurology at the University of Virginia, and as an adjunct associate professor at the University of Georgia and the Emory University School of Medicine. Most of Macciocchi’s research has been in the area of brain injury, while his clinical work has involved a more diverse subject matter.

Macciocchi performed a neuropsychological evaluation of Jefferson, which consisted of various tests designed to examine intellectual functioning, problem

37

solving, executive skills, tension concentration, language, and memory. He also reviewed Jefferson’s medical and school records, affidavits from Jefferson’s family, and Merikangas’s, Price’s, and Dudley’s expert reports, and conducted an interview with Jefferson. Macciocchi’s testing determined that Jefferson has an average level of IQ of 93, but also found a significant discrepancy in Jefferson’s verbal and performance IQ scores. However, unlike Drs. Dudley, Price, and Merikangas, Macciocchi said that this discrepancy was not indicative of brain injury, though he acknowledged that in the late 1980s, experts would have interpreted that result as a sign of brain dysfunction. The testing also revealed, according to Macciocchi, that Jefferson was between low-average and average in terms of executive functioning. Many of Macciocchi’s other tests revealed average or low-average performance.

After interviewing Jefferson, Macciocchi concluded that Jefferson did not exhibit any behavior to support a diagnosis of frontal lobe disorder. Macciocchi agreed that Jefferson likely has a learning disability, but did not agree that he consequently suffers from brain damage. Macciocchi testified that Jefferson’s injury at age two did not lead to the kind of symptoms associated with moderate- to-severe brain injuries that result in lifelong behavioral issues. Although Macciocchi could not rule out the possibility of brain injury, he opined that it was unlikely that Jefferson suffered this kind of injury.

38

Drs. Merikangas and Price heavily critiqued Dr. Macciocchi’s report. For one thing, Dr. Merikangas said that Macciocchi was not qualified to conduct neurological testing because he was neither a neurologist nor a medical doctor. Merikangas disputed Macciocchi’s opinion that Jefferson’s ability to conform his conduct to the requirements of structured environments suggested that he did not suffer from brain damage; according to Dr. Merikangas, people with brain damage usually do well in structured environments, and only have problems when dealing with more chaotic, unstructured situations. Merikangas discounted Macciocchi’s and Cella’s testimony that their interactions with Jefferson were normal as being largely irrelevant because people with brain injuries may be perfectly capable of controlling their behavior under most circumstances.

Nor was Merikangas impressed with the observation that the car accident had not fractured Jefferson’s skull, since, he opined, many individuals suffer brain injuries from a forceful impact on the skull, even when the skull does not fracture. As for Macciocchi’s suggestion that the relatively short hospital stay after the car accident meant the injury was not severe, Merikangas countered that five days is a long hospital stay for a child and indicated that serious complications were involved. Merikangas maintained that Macciocchi’s report did not make him question any of his conclusions.

39

Dr. Price likewise found nothing in Macciocchi’s report that caused him to question his diagnosis or opinions. Price pointed out that in scoring Jefferson on various psychological and intelligence tests, Macciocchi had used “demographically corrected” norms, which are norms that take into account factors like race, age, education, and gender in assessing a person’s test score. This is in contrast to the methodology employed by Dr. Price, who used “absolute” norms provided by the test manufacturers, which are standardized norms that don’t take into account demographic characteristics in scoring. Notably, the norms used by Macciocchi would not have been available at the time of Jefferson’s trial because they were not published until after 1991. These norms modify a person’s score in order to incorporate things like education level, race, age, and gender in light of how adults who have developed normally and who possess particular demographic traits score.

Dr. Price explained that because “demographically corrected normative standards are based upon performances of adults who have developed normally . . . and have . . . no known history of brain injury or disease[,] [i]t follows logically that such norms should be used with great caution, if at all, to identify acquired brain dysfunction in patients who have developmental disorders.” Price’s comment about the limited efficacy of using norms as a diagnostic tool came from the doctors who had designed the test in the first place. Thus, according to Price,

40

Macciocchi’s use of the demographically corrected norms on Jefferson, someone who suffered a known head injury in early childhood, reflected an improper method of assessing whether Jefferson suffers from organic brain dysfunction.

3.

Following the evidentiary hearings, the district court issued an opinion determining, first, that the state habeas court had not afforded Jefferson a full, fair, and adequate hearing consistent with due process, and, thus, that the state habeas court’s findings were not entitled to a presumption of correctness. Jefferson v. Sellers, 250 F. Supp. 3d 1340, 1354 (N.D. Ga. 2017). After developing the record and carrying out the fact-finding set forth in the Supreme Court’s remand order, the district court applied the prior version of 28 U.S.C. § 2254(d)(2), (6) and (7) -- the statutory provisions the Supreme Court deemed particularly relevant to Jefferson’s case. The district court concluded that the fact-finding procedure employed by the state habeas corpus court was fundamentally unfair, and, therefore, declined to apply the presumption of correctness to its findings of fact.

Once the district court determined that the state court’s findings of fact did not warrant any deference, it turned to the task of resolving the factual disputes involved in Jefferson’s ineffective-assistance-of-counsel claim. First, the court found that Dr. Dudley’s statement that he never backed away from his written recommendation that additional neuropsychological testing would be worthwhile

41

was more credible than Schuster’s claim that Dudley orally had dismissed further testing because it would be a waste of time. In making this finding, the district court cited to the demeanors of Schuster and Dudley during the evidentiary hearings, the absence in Schuster’s billing records of any notation that he spoke to Dudley by phone, Cella’s testimony contradicting Schuster’s about whether they talked to Dudley, and inconsistencies between Schuster’s testimony at the state habeas court evidentiary hearings and his testimony at the district court’s evidentiary hearings.

Second, the district court found Jefferson’s experts to be more credible and convincing than the State’s rebuttal expert. The court expressed considerable skepticism about Dr. Macciocchi’s testimony because, among other things, he had no medical expertise in neurology and because he used improper norms in scoring Jefferson’s test results. The district court also emphasized that “Macciocchi’s report did not in any way cause Dr. Merikangas or Dr. Price to doubt the conclusions they reached.” As a result, the district court concluded that Jefferson had presented “a compelling case that he suffers (and suffered) from organic brain damage.”

Based on these findings, the district court determined that Schuster and Cella had provided deficient performance, placing great emphasis on the fact that counsel had ignored an “unqualified recommendation” from their retained

42

psychologist that further neuropsychological testing should be done. The district court also concluded that counsel’s actions had prejudiced Jefferson, emphasizing the compelling evidence of organic brain damage that likely would have changed how the sentencing jury saw Jefferson’s moral culpability, would have weakened the aggravating factors the jury relied on, and would have been largely consistent with the defense strategy actually pursued at the sentencing phase.5 The court vacated Jefferson’s death sentence and ordered a new sentencing proceeding.

The Warden has timely appealed from the district court’s order, arguing that the court erred in both determining that no presumption of correctness was warranted, and in finding Jefferson’s trial counsel constitutionally ineffective.

III.

Because Jefferson filed his habeas corpus petition on April 23, 1996, one day before AEDPA’s effective date, we review his claims under the standards

5 The district court summarized its conclusion this way:

The bottom line is that Petitioner had his head run over by a car when he was only two years of age, and he suffered a traumatic injury to his head as a result. While the opinions of Petitioner’s experts regarding the impact of this traumatic injury had on Petitioner are not due to be accepted without critique or careful consideration, especially in light of the challenges Respondent has raised, a jury should have had an opportunity to hear and consider those opinions. Due to trial counsel’s deficient performance, Petitioner’s sentencing jury had no such opportunity. Based on this Court’s careful review of the mental health evidence presented by both Petitioner and Respondent, Petitioner has satisfied the Court that there is a reasonable probability that at least one juror would have found the mental health evidence presented compelling enough to vote against death . . . .

Jefferson, 250 F. Supp. 2d at 1383-84 (footnote omitted).

43

found in the pre-AEDPA statute. Bradshaw v. Stumpf, 545 U.S. 175, 182 (2005); Lindh v. Murphy, 521 U.S. 320, 323 (1997). Questions of law and mixed questions of law and fact resolved by state habeas courts are reviewed de novo, while factual findings made by the state habeas courts are presumed to be correct unless one of eight exceptions enumerated in § 2254(d) applies. 6 Freund v.

6 The pre-AEDPA version of Section 2254(d) read this way:

(d) In any proceeding instituted in a Federal court by an application for a writ of habeas corpus by a person in custody pursuant to the judgment of a State court, a determination after a hearing on the merits of a factual issue, made by a State court of competent jurisdiction in a proceeding to which the applicant for the writ and the State or an officer or agent thereof were parties, evidenced by a written finding, written opinion, or other reliable and adequate written indicia, shall be presumed to be correct, unless the applicant shall establish or it shall otherwise appear, or the respondent shall admit --

(1) that the merits of the factual dispute were not resolved in the State court hearing;

(2) that the factfinding procedure employed by the State court was not adequate to afford a full and fair hearing;

(3) that the material facts were not adequately developed at the State court hearing;

(4) that the State court lacked jurisdiction of the subject matter or over the person of the applicant in the State court proceeding;

(5) that the applicant was an indigent and the State court, in deprivation of his constitutional right, failed to appoint counsel to represent him in the State court proceeding;

(6) that the applicant did not receive a full, fair, and adequate hearing in the State court proceeding; or

(7) that the applicant was otherwise denied due process of law in the State court proceeding;

(8) or unless that part of the record of the State court proceeding in which the determination of such factual issue was made, pertinent to a determination of

44

Butterworth, 165 F.3d 839, 861 (11th Cir. 1999) (en banc). If one of the exceptions is found, the state court’s fact-finding is stripped of a presumption of correctness. Kelley v. Sec’y for Dep’t of Corr., 377 F.3d 1317, 1335 (11th Cir. 2004) (citations omitted). The petitioner then must establish “the facts necessary to support his claim by only a preponderance of the evidence.” Id.

Findings of fact made by the district court are reviewed for clear error.

Williams v. Turpin, 87 F.3d 1204, 1209 (11th Cir. 1996). A factual finding may be reversed for clear error only when the Court is “left with a definite and firm conviction that a mistake has been committed.” United States v. Rothenberg, 610 F.3d 621, 624 (11th Cir. 2010). “If the district court’s findings of fact are plausible in light of the record viewed in its entirety, the court of appeals must accept them even if it is convinced that had it been sitting as the trier of fact, it would have

the sufficiency of the evidence to support such factual determination, is produced as provided for hereinafter, and the Federal court on a consideration of such part of the record as a whole concludes that such factual determination is not fairly supported by the record:

And in an evidentiary hearing in the proceeding in the Federal court, when due proof of such factual determination has been made, unless the existence of one or more of the circumstances respectively set forth in paragraphs numbered (1)

to (7), inclusive, is shown by the applicant, otherwise appears, or is admitted by the respondent, or unless the court concludes pursuant to the provisions of paragraph numbered (8) that the record in the State court proceeding, considered as a whole, does not fairly support such factual determination, the burden shall rest upon the applicant to establish by convincing evidence that the factual determination by the State court was erroneous.

28 U.S.C. § 2254 (emphases added).

45

weighed the evidence differently.” United States v. Fidelity Capital Corp., 920 F.2d 827, 836 n. 36 (11th Cir. 1991) (quotation omitted).

“[B]oth the performance and prejudice components of the [ineffective assistance of counsel] inquiry are mixed questions of law and fact,” and we review them de novo. Strickland v. Washington, 466 U.S. 668, 698 (1984); DeBruce v. Comm’r, Alabama Dep’t of Corr., 758 F.3d 1263, 1266 (11th Cir. 2014). We also review the reasonableness of a strategic decision de novo, Provenzano v. Singletary, 148 F.3d 1327, 1330 (11th Cir. 1998), while the antecedent inquiry of whether the action was in fact a strategic one, rather than the result of neglect, is a question of fact, and the district court’s determination in that respect is presumed to be correct unless it’s clearly erroneous, Horton v. Zant, 941 F.2d 1449, 1462 (11th Cir. 1991).

IV.

A.

Our first task, then, is to determine whether the state habeas court’s factual findings are entitled to a presumption of correctness. We hold they are not.

The district court conducted an extensive hearing to address the essential question about the state habeas court proceedings posed by the Supreme Court, and thereafter, made detailed findings. It began by finding that the order adopted by Judge Newton was drafted exclusively by the State pursuant to an ex parte request

46

made by Newton’s law clerk to Assistant Attorney General Smith. Jefferson, 250 F. Supp. 3d at 1351. Moreover, the district court found that the state habeas court did not provide any drafting instructions or guidance as to his own findings and conclusions, other than simply telling the Assistant Attorney General at the highest order of generality that the proposed order should deny habeas relief. Id. The district court also found that Judge Newton did not ask the Petitioner to submit his own proposed factual findings and conclusions of law, and that the state judge did not so much as inform the Petitioner of the request he had made of the Assistant Attorney General. Id.

The district court also observed that a comparison between Assistant Attorney General Smith’s proposed order and the order ultimately adopted by the state habeas court shows that Judge Newton “uncritically accepted findings prepared without judicial guidance.” The state court order referenced an attorney’s affidavit that was never submitted in the case and the order contained at least twenty-one typographical errors that Judge Newton either never saw or overlooked. Id. (quotation omitted). The district court expressly found that: “These errors and issues invited the reasonable inferences that Judge Newton failed to review the proposed order submitted by Ms. Smith before executing the final order and failed to review the relevant evidence.” Id.

47

We can find no clear error in the district court’s fact findings,7 and we agree that Jefferson was deprived of a full and fair hearing under § 2254(d). The version of § 2254(d) at play in this case dates back to 1966, when Congress “enacted an amendment to the federal habeas statute [28 U.S.C. § 2254] that was an almost verbatim codification of the standards delineated in Townsend v. Sain[, 372 U.S. 293 (1963), overruled on other grounds by Keeney v. Tamayo–Reyes, 504 U.S. 1 (1992)].” Jefferson, 560 U.S. at 290. Townsend, for its part, had addressed the proper standards for federal courts to apply to evidentiary hearings in federal habeas corpus cases. It held that “[w]here the facts are in dispute, the federal court in habeas corpus must hold an evidentiary hearing if the habeas applicant did not receive a full and fair evidentiary hearing in a state court, either at the time of the trial or in a collateral proceeding.” Townsend, 372 U.S. at 312. It then laid out various factors to consider when deciding whether a petitioner had received “a full and fair evidentiary hearing in a state court,” id. at 313, most of which appeared word-for-word in Congress’s subsequent codification in § 2254(d). 28 U.S.C. § 2254(d) (1966).

Sometime later, in Anderson v. Bessemer City, 470 U.S. 564 (1985), the Supreme Court considered, in a Title VII case, the significance of a trial court’s

7 The State takes issue with a handful of the district court’s findings, but we are unimpressed. Most of the State’s arguments simply quibble with the district court’s characterization of the record, and fail to point to any clear error in the findings.

48

adoption of one party’s proposed findings of fact and conclusions of law. Id. at 572. While criticizing the practice generally, the Supreme Court held that the district court’s adopted proposed order in that case did not require more stringent appellate review because: (1) the court itself had provided the framework for the proposed findings when it issued its preliminary memorandum, which set forth its essential findings and directed petitioner’s counsel to submit a more detailed set of findings consistent with them; (2) the respondent had been provided and availed itself of the opportunity to respond at length to the proposed findings; and (3) the court did not “simply adopt petitioner’s proposed findings: the findings it ultimately issued -- and particularly the crucial findings regarding petitioner’s qualifications, the questioning to which petitioner was subjected, and bias on the part of the committeemen -- vary considerably in organization and content from those submitted by petitioner’s counsel.” Id. at 572-573.

Notably, Judge Newton did not provide a preliminary memorandum outlining his essential findings of fact or conclusions of law. Nor did he state anything in open court about how he viewed the case. Indeed, he did not provide any guidance to the Assistant Attorney General about how the proposed order was to be drafted or what was to be included, except simply telling the State that the proposed order should deny habeas relief. Moreover, the state habeas judge did not afford Jefferson any opportunity to respond to Smith’s proposed order, nor did

49

he inform Jefferson that a request had been made of the Assistant Attorney General to prepare proposed findings of fact and conclusions of law. Even after Petitioner learned from Smith of the ex parte request, Judge Newton still did not provide him an opportunity to offer criticism of the proposed order or to submit his own proposed order. Finally, other than altering the concluding sentence and including the date and his signature, Judge Newton made no changes to the proposed order, notwithstanding the many glaring errors it contained.

Neither this Court nor the Supreme Court has ever reviewed the consequence of an adopted verbatim opinion in the context of the prior version of §§ 2254(d)(2), (6), and (7) (1966) -- which is exactly what the Supreme Court directed us to do here. 8 In the remand order, the Supreme Court determined that our earlier opinion had erred by “treating § 2254(d)(8) as the exclusive statutory exception, and by failing to address Jefferson’s argument that the state court’s procedures deprived its findings of deference.” Jefferson, 560 U.S. at 292-93 (emphasis omitted). It instructed us to address whether the state court’s “factfinding procedure,” “hearing,” and “proceeding” were “full, fair, and adequate,” under other provisions of the statute, citing §§ 2254(d)(2), (6), (7)

8 The closest we’ve come was when we reviewed whether a verbatim adopted order violated § 2254(d)(6), in Brownlee v. Haley, 306 F.3d 1043 (11th Cir. 2002), but there we said that “the extensive record from the Rule 32 proceedings in this case ‘eliminates any doubt about the [trial] judge’s involvement in the matter and careful analysis of . . . [the] evidence.’” Id. at 1067 n.19 (citation omitted). Thus, Brownlee did not raise the same concerns that the Supreme Court flagged in this case, nor can we find any assurances here like the ones we cited to in Brownlee.

50

(1966). Id. It suggested that based on the facts -- that the proposed order was solicited ex parte, Jefferson was not given an opportunity to contest the proposed order or offer an alternative one, and internal evidence in the order suggested the state habeas judge did not review it -- the state habeas court might not have afforded Jefferson fair process, which would strip from its factual determinations the presumption of correctness under § 2254(d) (1966). Id. at 294.

To the extent the State claims the district court construed the pre-AEDPA § 2254(d) too broadly, we disagree. The district court considered § 2254(d)(2), (6), (7), individually and in concert, just as the Supreme Court directed on remand and as we’ve done before. See Stokes v. Singletary, 952 F.2d 1567, 1573-74 (11th Cir. 1992); Wesley v. Alabama, 488 F.2d 30, 31-32 (5th Cir. 1974); Lane v. Henderson, 480 F.2d 544, 545 (5th Cir. 1973).

As for the State’s argument that the Georgia Supreme Court’s review of and deference to the state habeas trial court’s findings cleared up any procedural irregularities that may have inhered in the trial court’s findings in the first instance, the Georgia Supreme Court’s decision does not contain any findings of fact separate from those made by the state trial court that are entitled to § 2254(d) (1966)’s presumption of correctness. For starters, the critical finding -- whether Jefferson’s trial counsel or his psychologist was more credible -- is not the kind of factual finding entitled to the presumption when made by an appellate court. See

51

Cabana v. Bullock, 474 U.S. 376, 388 n.5 (1986) (noting that the presumption would not apply to “credibility determinations that could not be accurately made by an appellate court on the basis of a paper record”). Moreover, the Georgia Supreme Court did not engage in any new factfinding in Jefferson’s appeal, but instead gave deference to the state habeas trial court’s fact findings under Georgia state law. See Jefferson, 431 S.E.2d at 112 (applying Ga. Code Ann. § 9-11-52, which provides that “[f]indings shall not be set aside unless clearly erroneous, and due regard shall be given to the opportunity of the trial court to judge the credibility of the witnesses”). In any event, the United States Supreme Court’s remand order makes clear that in resolving “the legal status of the state court’s factual findings,” it was concerned about the procedure used, not just the result reached, and the Georgia Supreme Court’s review did not remove the taint of procedural irregularity involved in the state habeas trial court’s fact finding.

Taking all of this together, we agree with the district court that the factual record compels the conclusion that Jefferson did not receive a full and fair hearing in his state court habeas proceeding. We, therefore, hold that the state court’s factual findings were not entitled to a presumption of correctness under pre- AEDPA law.

We neither hold nor imply that if this were a post-AEDPA case the state court’s adoption of the order in these same circumstances would affect the

52

presumption of correctness afforded by 28 U.S.C. § 2254(e)(1) (1996) or the deference provided factual determinations by § 2254(d)(2) (1996). See Jones v. GDCP Warden, 753 F.3d 1171, 1183 (11th Cir. 2014) (explaining that the Supreme Court’s decision in Jefferson “never could have held, nor did it presume to hold, that this kind of adopted order is not entitled to AEDPA deference”); see also Rhode v. Hall, 582 F.3d 1273, 1281–82 (11th Cir. 2009) (Where both sides had opportunity to present the state habeas court with their version of the facts, the court’s verbatim adoption of the State’s proposed order, which petitioner contended selectively used and mischaracterized evidence, did not deprive the findings of AEDPA deference.).

B.

We come then to the substantive question underlying this case: whether, based on new evidence submitted at the federal hearing, along with all of the other evidence adduced earlier, Jefferson received ineffective assistance of counsel in violation of the Sixth Amendment at the sentencing phase of his trial. The Sixth Amendment entitles a criminal defendant to a lawyer, and that right includes the right to effective counsel. McMann v. Richardson, 397 U.S. 759, 771 n.14 (1970). Claims of ineffective assistance of counsel are governed by the well-established principles found in Strickland v. Washington, 466 U.S. 668 (1984). To prevail under Strickland, a petitioner must make two showings: first, he must establish that

53

counsel’s representation “fell below an objective standard of reasonableness”; and second, “that there is a reasonable probability that, but for counsel’s unprofessional errors, the result of the proceeding would have been different.” Darden v. Wainwright, 477 U.S. 168, 184 (1986) (quotations omitted). “A reasonable probability is a probability sufficient to undermine confidence in the outcome.” Strickland, 466 U.S. at 694.

We are especially cautious in evaluating ineffective assistance of counsel claims since “it is all too easy to conclude that a particular act or omission of counsel was unreasonable in the harsh light of hindsight.” Bell v. Cone, 535 U.S. 685, 702 (2002). We therefore evaluate ineffective assistance claims from the perspective of the time of trial because “a more rigid requirement ‘could dampen the ardor and impair the independence of defense counsel, discourage the acceptance of assigned cases, and undermine the trust between attorney and client.’” Lockhart v. Fretwell, 506 U.S. 364, 372 (1993) (quoting Strickland, 466 U.S. at 690).

Nevertheless, where a petitioner has established that his lawyer’s actions fell below an objectively reasonable standard of performance and that such ineffectiveness rendered his trial fundamentally unfair, he is entitled to relief even in spite of the substantial deference that we show to the decisions of trial counsel. See Chandler v. United States, 218 F.3d 1305, 1314 (11th Cir. 2000) (en banc).

54

On the record now before us, we are satisfied that Jefferson has made that showing.

1.

A lawyer’s duty to render effective assistance of counsel necessarily includes the duty to undertake appropriate investigations in support of the decisions he makes and the actions he takes at trial. Strickland, 466 U.S. at 691 (“[C]ounsel has a duty to make reasonable investigations.”). As the Supreme Court has explained, “strategic choices made after thorough investigation of law and facts relevant to plausible options are virtually unchallengeable; and strategic choices made after less than complete investigation are reasonable precisely to the extent that reasonable professional judgments support the limitations on investigation.” Id. at 690-91. In other words, when a limited investigation was performed, our inquiry under Strickland’s first prong focuses on the reasonableness of the decision to curtail investigative efforts. See Grayson v. Thompson, 257 F.3d 1194, 1225 (11th Cir. 2001) (“[E]ven when trial counsel’s investigation and presentation is less complete than collateral counsel’s, trial counsel has not performed deficiently when a reasonable lawyer could have decided, under the circumstances, not to investigate or present particular evidence.”). Where the decision to place limits on an investigation represents a strategic choice, it is “accorded a strong presumption of reasonableness.” Mills v.

55

Singletary, 63 F.3d 999, 1021 (11th Cir. 1995). If, however, the failure to engage in further investigation was the result of neglect or oversight, no presumption of reasonableness attaches, and we are far more likely to find deficient performance. See Baxter v. Thomas, 45 F.3d 1501, 1513 (11th Cir. 1995); Middleton v. Dugger, 849 F.2d 491, 493-94 (11th Cir. 1988); Harris v. Dugger, 874 F.2d 756, 763 (11th Cir. 1989); Armstrong v. Dugger, 833 F.2d 1430, 1433 (11th Cir. 1987).

We now hold that Jefferson’s lawyers’ performance fell beneath an objective standard of reasonableness. The outcome of Jefferson’s appeal is different the second time around. The essential facts as found by the district court on remand establish that Jefferson’s trial counsel inexplicably ignored the clear written recommendation of their retained psychologist to seek further neuropsychological evaluation in order to rule out an organic etiology.

As we’ve recounted, Schuster testified before the district court that he spoke with Dudley after Dudley had evaluated Jefferson and, on the basis of that discussion, he concluded that a neuropsychological evaluation of Jefferson would not be worthwhile. Schuster claimed that Dudley told him that Jefferson “was a sociopath or a psychopath and not to worry about it,” and that “Mr. Jefferson was the way Mr. Jefferson was.” Schuster also said that he believed the line in Dudley’s written report recommending further testing was just a “throw-out in the report.”

56

Dudley, however, told a very different story. According to the doctor’s testimony in federal court, he never backed off his written recommendation for further testing that he included in his written evaluation of Jefferson. He denied ever telling Schuster that further testing was unnecessary and also denied ever saying that Jefferson was “just a criminal.” Dudley explained that at the time he submitted his report to Schuster and Cella he “assumed that additional testing would be done.” Dudley’s testimony was also consistent with his affidavit filed in the state habeas court, which denied telling Jefferson’s counsel that a neuropsychological evaluation would not be necessary or that it would not be worthwhile.

The district court weighed the testimony, found Dudley more credible, and chose to accept his account of what happened. Its decision to do so was not clearly erroneous. A district court’s plausible weighing of contradictory testimony will not be second-guessed on appeal. See United States v. Garcia, 890 F.2d 355, 359 (11th Cir. 1989); Childrey v. Bennett, 997 F.2d 830, 834 (11th Cir. 1993). We are therefore presented with these facts: Jefferson’s counsel retained a psychologist who examined the Petitioner and recommended that neuropsychological testing be conducted to determine whether he suffered from organic brain damage as a result of his head being run over by a car when he was two years old. Yet his lawyers ignored this unambiguous recommendation and failed to obtain any additional

57

testing. The injury was so severe he was hospitalized for five days, left with a cranial indentation and an abnormally enlarged skull, and suffered frontal lobe and neurological damage.

Without any explanation for counsel’s failure to pursue further testing, the district court found that trial counsel acted out of neglect rather than as the product of strategic judgement. We cannot say that this finding is clearly erroneous. For one thing, without Schuster’s testimony, the State can point to no strategic reason for why Dudley’s recommendation was ignored. Worse still, on remand the district court heard testimony from Jefferson’s other attorney, Cella, who claimed that the actual reason why no neuropsychological testing had been scheduled was because of inexperience on counsel’s part. Cella admitted that “I should have followed up on [Dudley’s recommendation]. And had I had more -- if this had not been my first death penalty case, I expect I would have.”

In the past, we’ve held that the failure to engage in further investigation was the result of neglect -- rather than a strategic decision -- on very similar facts. In DeBruce, trial counsel obtained a competency report detailing the defendant’s difficult and deprived upbringing that contradicted his mother’s account of his childhood as being healthy and uneventful. 758 F.3d at 1274. Rather than investigating the claims made in the report, trial counsel ignored it. Id. at 1275. In determining that counsel acted out of neglect rather than on account of any tactical

58

consideration, we explained that “counsel’s testimony during the state collateral proceedings made clear that his failure to conduct adequate investigation was not the result of a strategic decision, but rather a result of the lack of time as he scrambled to prepare for [petitioner’s] capital trial.” Id. Similarly, in Harris v. Dugger, 874 F.2d 756 (11th Cir. 1989), rather than offering a strategic justification for their failure to investigate the defendant’s background, “[e]ach lawyer testified that he believed that the other was responsible for preparing the penalty phase of [the] case,” meaning that neither attorney looked into the defendant’s background for mitigating evidence. Id. at 763. Under these circumstances we determined that trial counsel’s ignorance of crucial evidence “resulted not from an informed judgment, but from neglect,” and, therefore, we did not apply the presumption of reasonableness that would otherwise attach to strategic decisions. Id. And in Armstrong v. Dugger, 833 F.2d 1430 (11th Cir. 1987), a panel of this Court made a similar determination where trial counsel “explicitly testified that [he did not make] in any way a strategic or tactical decision” to limit investigation but rather that his failure to investigate was a result of inexperience. Id. at 1433.

The State argues nevertheless that defense counsel’s actions should be regarded as strategic because they pursued a residual doubt defense at the penalty phase, and thus had no need for neuropsychological evidence to mount their case. However, the penalty phase transcript shows that Jefferson’s lawyers did not

59

emphasize residual doubt at sentencing, indeed they only barely mentioned it in passing; rather they focused on Jefferson’s hard upbringing, the need to show Jefferson mercy, and the idea that Jefferson was not fully responsible for his actions. The following excerpts from Schuster’s and Cella’s arguments give an accurate flavor of what arguments they made:

“Jefferson was a poor black kid who grew up with a lot of other poor black kids who didn’t have anywhere to go, had very few positive influences in their lives.”

…

“We could go down into Atlanta and pick out at random any black 17-yearold we wanted to, and we’d get an arrest record very similar to Lawrence Jefferson’s.”

…

“[Jefferson] doesn’t have [a] history of violence. He has a good work record.

. . . He is ambitious. Came all the way down here to find work to support his family better than he was able to up there in Louisville. That shows ambition. It shows a sense of caring for people that he loves.”

…

“You know, Lawrence’s background, his life experience, has caused him to strike out at a defenseless person, such as Ed Taulbee, that he killed without necessity, and y’all found that. But aren’t we as a society better than that?”

…

“We are asking you, please, look in your hearts, find some understanding, some love, some compassion for Lawrence.”

It’s true that defense counsel also made a few brief allusions to residual doubt at Jefferson’s sentencing. For instance, Schuster made the remark that “you [may] find a man guilty beyond a reasonable doubt, but that’s not all doubt.” Counsel also noted that there were no eye witnesses to the crime, no guilty plea,

60

and “no evidence of any consciousness of guilt.” However, these remarks were few and far between, and did not form the foundation of the closing arguments.

Instead, the argument made at Jefferson’s sentencing hearing is comparable to the “halfhearted mitigation case” put forward in Wiggins v. Smith, 539 U.S. 510 (2003), which “did not focus exclusively on [the defendant’s] direct responsibility” but rather also involved references to the defendant’s difficult upbringing, albeit without fully developing the facts relevant to a plea for mercy. Id. at 526. On those facts, the Supreme Court rejected the argument that defense counsel had made a strategic decision to focus on direct responsibility, and instead reasoned that “the ‘strategic decision’ [] respondents . . . invoke to justify counsel’s limited pursuit of mitigating evidence resembles more a post hoc rationalization of counsel’s conduct than an accurate description of their deliberations prior to sentencing.” Id. at 526-27.

The same can be said here. Invoking a residual doubt strategy in order to justify the decision to forgo further neuropsychological testing as a strategic choice is unconvincing since counsel can barely be said to have employed a residual doubt defense. And what’s more, as we lay out later, the very evidence they failed to pursue would have powerfully bolstered the limited mitigation case they did present, to explain the contradiction between Jefferson’s background and the murder, and the compassion they sought from the jury.

61

We recognize that deficient performance is an objective inquiry, meaning that even if trial counsel acted out of inexperience or neglect, his actions would still not constitute ineffective assistance of counsel if we can say that a reasonable attorney could have acted as he did. Gordon v. United States, 518 F.3d 1291, 1301-02 (11th Cir. 2008) (“Because [the Strickland] standard is objective, it matters not whether the challenged actions of counsel were the product of a deliberate strategy or mere oversight. The relevant question is not what actually motivated counsel, but what reasonably could have motivated counsel.”). Thus, the subjective failings of Cella and Schuster are not dispositive. Jefferson still must show that no reasonable attorney would have behaved as defense counsel did in this case. Chandler, 218 F.3d at 1315 (“[A] petitioner must establish that no competent counsel would have taken the action that his counsel did take.”).

He has done so here. Importantly, the Supreme Court has explained that a “failure to investigate thoroughly [that stems] from inattention, not strategic judgment,” serves to “underscore[] the unreasonableness of counsel’s conduct.” Wiggins, 539 U.S. at 526. The failure of Jefferson’s trial counsel to arrange for neuropsychological testing resulted in the exact kind of non-strategic decision that underscores the unreasonableness of their conduct. As we’ve said many times, “[a] failure to investigate can be deficient performance in a capital case . . . when counsel ignores ‘red flags’ that any reasonable attorney would perceive as

62

demanding further investigation.” Housel v. Head, 238 F.3d 1289, 1294 (11th Cir. 2001). So, in Cunningham v. Zant, 928 F.2d 1006 (11th Cir. 1991), we found deficient performance where trial counsel was aware that the defendant had suffered a near-fatal head injury but failed to seek any expert assessment of the potentially mitigating evidence associated with the injury. Id. at 1018. Similarly, in DeBruce, this Court found deficient performance where trial counsel failed to follow up on information contained in a competency report about the defendant where the report indicated that the defendant’s background was deeply troubled. 758 F.3d at 1274. As the Court explained, trial counsel “unreasonably ignored” the report, particularly since its contents contradicted other evidence the defense lawyer had gathered, which would have put any reasonable attorney on notice that more investigation was needed. Id. at 1275. Again in Williams v. Allen, 542 F.3d 1326, 1340 (11th Cir. 2008), we found deficient performance where “the information that trial counsel did acquire would have led a reasonable attorney to investigate further” but where trial counsel failed to do so. Id. at 1340. In particular, the Court found that no reasonable attorney would have curtailed his investigation into the defendant’s troubled background where both a PSI and a report from a retained psychologist indicated that the defendant had a history of mental health problems. Id. “To the contrary, the many red flags . . . would have a prompted a reasonable attorney to conduct additional investigation.” Id.

63

Here too the particular circumstances were of the kind that would’ve impelled any reasonable lawyer toward further, more robust investigation. Cella and Schuster knew Jefferson had suffered a serious head injury as a child. He had a visible, prominent scar across his skull and his head was asymmetrical. Counsel knew Jefferson had had an unusual interaction with Dudley, when he first refused to cooperate with him and later explained, “I get that way sometimes, man.” Counsel also knew that when he was arrested, Jefferson acted strangely, telling police that he didn’t need to be around other people, that he wanted to be executed, and that he wanted to be put to sleep. In fact, Cella testified at the state habeas hearing that he knew, before trial, that Jefferson had admitted during his interrogation by police officers that:

I don’t need to be around people . . . . I talk to myself. I ain’t crazy. I don’t know nothing. I got hit by an auto when I was a kid. I had another dream and saw the man in my dream. I call myself Tony. I don’t think I am two people. How quick can I be put to sleep? I don’t feel anything about this. Do you believe in Heaven and Hell? . . . I want to be put to sleep . . . . I want to be executed.

Jefferson, 570 F.3d at 1290. And, most importantly, counsel had received Dudley’s written report, which made a crystal clear, uncontroverted recommendation that further neuropsychological testing would be worthwhile. To all these red flags, defense counsel turned a blind eye, without citing anything remotely suggesting that further testing would have impaired their case.

64

These facts make Jefferson’s case noticeably different from other cases where a failure to engage in further investigation, even where red flags were present, did not render counsel ineffective. In Housel v. Head, 238 F.3d 1289 (11th Cir. 2001), for instance, this Court did not find deficient performance even though trial counsel failed to pursue additional mental health investigation and knew that the defendant had suffered a serious head injury in a car accident. Id. at 1296. There, defense counsel had not observed any unusual behavior in their interactions with the defendant, and they had already had two psychological evaluations conducted, “neither of which offered any encouragement to proceed further.” Id. Here, in sharp contrast, the psychologist who evaluated Jefferson expressly recommended further testing, and did so in writing, and while Jefferson’s lawyers had not personally observed any unusual behavior from Jefferson, they were aware that Dr. Dudley and the investigating police officers had.

Moreover, even if trial counsel “prepared extensively” for the sentencing phase in other ways -- by, for instance, traveling to Louisville, interviewing many of Jefferson’s family members and acquaintances, and consulting with death penalty specialists from the ACLU and the Southern Poverty Law Center -- it is still clear that the decision to forgo neuropsychological testing was an unreasonable one and in no way justified by the other unrelated steps they took. The Supreme Court has held that “defense counsel’s efforts to find mitigating

65

evidence by other means [does not] excuse[] them” from inquiring into evidence highly relevant to sentencing that “[n]o reasonable lawyer would forgo” examining. Rompilla v. Beard, 545 U.S. 374, 388-90 (2005); see also Bobby v. Van Hook, 558 U.S. 4, 11 (2009) (explaining that while extensive investigation can sometimes excuse the decision to not investigate further, in other cases even a wide-ranging investigation will not be enough where the need for further investigation “would have been apparent from documents any reasonable attorney would have obtained”).

So, while the record before this Court the second time around is largely the same as it was a decade ago, it has changed in a few critical respects. Most importantly, the State can no longer offer any explanation for why Cella and Schuster failed to follow up on Dudley’s recommendation. The district court’s factual finding completely eliminates the claim that Dr. Dudley told Schuster it would be a waste of time to examine Jefferson any further, that Jefferson was just a criminal, and that Dudley would be of no help to Jefferson at sentencing. Absent any reasonable explanation, Cella and Schuster’s actions are not entitled to the presumption of reasonableness that attaches to strategic decisions. And without this presumption, it seems to us clear that no reasonable attorney would have failed to have Jefferson tested for organic brain damage when there were so many red flags calling out for further investigation and where the results of that inquiry in no

66

way undermined the defense they actually presented. Because the limits Cella and Schuster placed on their investigation were unreasonable, their actions at the penalty phase of Jefferson’s trial fell below the standard of constitutionally effective lawyering.

2.

Having found deficient performance, we turn to whether there is a reasonable probability that Jefferson suffered prejudice at the penalty phase of his trial. “In assessing prejudice, we reweigh the evidence in aggravation against the totality of available mitigating evidence.” Wiggins, 539 U.S. at 534. We weigh all of the good and all of the bad. “In that process, what matters is not merely the number of aggravating or mitigating factors, but their weight.” Reed v. Sec’y, Fla. Dep’t of Corr., 593 F.3d 1217, 1240-41 (11th Cir. 2010). “[W]e have held that a petitioner is prejudiced where the mitigation evidence omitted by counsel’s deficient investigation ‘paints a vastly different picture of the petitioner’s background than that created by’ the actual penalty-phase testimony.” DeBruce, 758 F.3d at 1276 (quoting Williams, 542 F.3d at 1342).

To put the doctrine in terms specific to Jefferson’s case, because Georgia requires that the jurors unanimously recommend death in order for the trial judge to impose the death penalty, Hill v. State, 301 S.E.2d 269, 270 (Ga. 1983), the relevant inquiry is whether “there is a reasonable probability that at least one juror

67

would have struck a different balance.” Wiggins, 539 U.S. at 537. “A reasonable probability is a probability sufficient to undermine confidence in the outcome.” Evans v. Sec’y, Dep’t of Corr., 703 F.3d 1316, 1326 (11th Cir. 2013). We conclude that there is a reasonable probability that at least one juror on Jefferson’s original sentencing jury would have struck a different balance and decided against imposing the death penalty had the jury been presented with the powerful evidence that Jefferson suffers from organic brain damage and its sequalae.

First, as we’ve mentioned, the evidence of brain damage that would have been introduced had Jefferson’s counsel performed in a constitutionally effective manner would have profoundly changed the character of the penalty phase of the proceedings by fundamentally transforming Jefferson’s sentencing profile. This sort of mitigation evidence is precisely the kind that may establish prejudice at the penalty phase. In DeBruce, for example, a panel of this Court found prejudice where trial counsel’s deficient performance gave the defendant’s jury “almost no reason to spare his life as demonstrated by the paucity of mitigating evidence actually presented during the penalty phase.” 758 F.3d at 1278. The Court noted that, as we explained then in finding prejudice, the sentencing jury had been presented with testimony that the defendant had had a somewhat impoverished but otherwise unremarkable childhood and had even briefly attended college, when the reality was that the defendant had struggled in school and dropped out at the age of

68

sixteen; he had been the recipient of severe, and nearly daily beatings from his siblings when he was younger; he grew up in a rough neighborhood where he witnessed the stabbing of one of his neighbors; he attempted to commit suicide on multiple occasions; and, finally, he had been treated for mental impairments. Id. at 1276. As the Court explained in finding prejudice, “adequate investigation would have enabled counsel to correct a positively misleading sentencing profile, and to present a vastly different picture than that created by the actual trial testimony.” Id. (quotation omitted).

In this case the sentencing profile actually presented at Jefferson’s trial compared with what would have been presented if Jefferson’s trial counsel had performed a proper investigation is striking. To illustrate, the thrust of the defendant’s closing arguments was that Jefferson had had a tough childhood; that although he was flawed, had an “attitude problem,” and was not a “good guy,” Jefferson was not an evil person and was capable of change; and that the jury should therefore have compassion and spare his life. Cella also invoked Jefferson’s hardscrabble background. Cella added that Jefferson’s upbringing made him no different than the many others who’ve grown up under difficult circumstances, saying that “Jefferson was a poor black kid who grew up with a lot of other poor black kids who didn’t have anywhere to go, had very few positive influences in their lives.” Similarly, Schuster reminded the jury that Jefferson had

69

grown up in poverty, and that although Jefferson was “responsible for his acts” and had “failed himself,” and the jury should show him mercy.

The evidence presented at the penalty phase also offered a fairly unremarkable portrait of Jefferson and his background. The only mention of his head injury at sentencing came from a fleeting remark his mother made about the incident, and neither Schuster nor Cella even mentioned it in their closing arguments. And Jefferson himself was not asked any questions about his mental health when he took the stand.

In sharp contrast to the arguments and evidence actually presented, the testimony of Drs. Price and Merikangas, and Jefferson’s mother and brother developed on collateral review paints a dramatically different portrait of who Jefferson is and what he was responsible for. In their telling, Jefferson suffers from organic brain damage. After the accident, Jefferson experienced severe, chronic headaches that made him cry, in addition to blackout spells that may have been petit mal seizures. The effects of this brain damage likely manifested in an inability to control his impulses and exercise sound judgment and meaningfully set him apart from an unimpaired person. It also makes it harder for Jefferson to deal with difficult or chaotic situations. Moreover, his mental impairments potentially resulted in “transient outbursts of rage” that are incongruous with Jefferson’s

70

ordinary behavior. In short, his organic brain damage “has had a profound and lifelong impact on Mr. Jefferson’s behavior.”

There is a powerful difference between someone who grew up poor and without a father and a person who grew up suffering from organic brain damage yielding debilitating mental impairments that worsened into adulthood. There is an even bigger difference between someone who has an “attitude problem” and someone whose frontal lobe was permanently damaged at a young age and who is therefore not capable of controlling his impulses or reactions to external stimuli at critical moments. The new mitigation evidence that almost surely would have been uncovered had Jefferson’s lawyers acted on Dr. Dudley’s recommendation not only would have bolstered the mitigation evidence already available but also would have dramatically transformed the sentencing profile presented to the jury. DeBruce, 758 F.3d at 1276; see also Brownlee, 306 F.3d at 1074.

Relatedly, the dramatically different sentencing profile created by the new mitigating evidence substantially weakens the aggravating factors relied on by the jury at sentencing. See Middleton v. Dugger, 849 F.2d 491, 495 (11th Cir. 1988) (“[P]sychiatric evidence . . . has the potential to totally change the evidentiary picture by altering the causal relationship that can exist between mental illness and homicidal behavior. Thus, psychiatric mitigating evidence not only can act in mitigation, it also could significantly weaken the aggravating factors.” (quotation

71

omitted)). A closely analogous example can be found in Ferrell v. Hall, 640 F.3d 1199 (11th Cir. 2011), where this Court found prejudice in substantial part because “evidence of [the defendant’s] mental illness measurably weaken[ed] the aggravating circumstances found by the jury.” Id. at 1234. In Ferrell, the sentencing jury relied on two aggravating factors -- that the murder was committed during the commission of a felony and was “outrageously and wantonly vile, horrible or inhuman in that it involved torture and depravity of mind.” Id. at 1235. In evaluating prejudice, this Court examined these factors in light of the mitigating evidence introduced on collateral review and determined that the aggravators were significantly undermined by the expert testimony that the defendant suffered from organic brain damage, which caused “increased impulsivity, decreased sound judgment, and [] actions that are not entirely volitional.” Id. As we explained, “the mental health expert opinions would have served to reduce the volitional nature of the crime, as well as Ferrell’s ability to plan and act rationally, and as a result, undercut the senselessness and cold-blooded nature of the crime as stressed by the prosecutor.” Id.

Nearly identical aggravating factors to those found in Ferrell were found by the jury that sentenced Jefferson -- that the offense of murder was committed while the defendant was engaged in the commission of another capital felony (i.e., armed robbery), Ga. Code Ann. § 17-10-30(b)(2) (1986); and that the offense of murder

72

“was outrageously or wantonly vile, horrible, or inhuman” in that it involved an aggravated battery to the victim, id. § 17-10-30(b)(7). See Ferrell, 640 F.3d at 1210. Jefferson’s expert witnesses likewise testified that Jefferson suffers from organic brain damage to the frontal lobe, which likely caused increased impulsivity, executive functioning problems, the inability to think with forethought, and unsound judgment. All of which would have provided for the first time a plausible explanation for an otherwise inexplicable and senseless crime. What’s more, this evidence also undermined the aggravator that the murder was committed in an “outrageously or wantonly vile, horrible, or inhuman” manner.

The mitigating evidence developed on collateral review also weakened the sentencing jury’s rationale for imposing the death penalty by reducing Jefferson’s moral culpability for the crime. As the Supreme Court has instructed, evidence of a troubled background and “diminished mental capacities” is “relevant to assessing a defendant’s moral culpability.” Wiggins, 539 U.S. at 535. Thus, for example, in Porter v. McCollum, 558 U.S. 30 (2009), the Court found prejudice, in part, because evidence of a “brain abnormality” not presented at sentencing would have allowed the jury to “accurately gauge [the defendant’s] moral culpability.” Id. at 41. Similarly, in Ferrell we found prejudice because the jury was never presented with evidence that the defendant “suffer[ed] from brain damage” affecting his “emotional stability” and “impulse control,” and the jury “labored under a

73

profoundly misleading picture of [the defendant’s] moral culpability because the most important mitigating circumstances were completely withheld from it.” Ferrell, 640 F.3d at 1236. And in DeBruce, we found prejudice where “passing mention of [the defendant’s] treatment for a mental disorder” meant that the jury did not hear “particularized characteristics of [the defendant’s] heavily disadvantaged background and upbringing” that would have been “relevant to assessing [the] defendant’s moral culpability.” DeBruce, 758 F.3d at 1276.

Here too the jury was deprived of extensive evidence tending to show that Jefferson suffers from brain damage that, significantly, made it difficult for him to control his behavior. The experts expressly testified that Jefferson is less morally culpable for his actions precisely because he is less able to control himself than would an unimpaired individual. The mitigation evidence developed on collateral review bears directly on Jefferson’s moral culpability and appreciably weakens whatever reasons the jury had for imposing a death sentence.

Moreover, there are no countervailing considerations in this case suggesting, as we’ve observed elsewhere, that the new mitigation evidence developed on collateral review would have exerted anything other than a salutary effect -- and a powerful one at that -- on the evidence actually presented and the arguments made by Jefferson’s lawyers. See generally Chandler, 218 F.3d at 1319 (“Considering the realities of the courtroom, more is not always better. Stacking defenses can

74

hurt a case.”); Collins v. Francis, 728 F.2d 1322, 1348 (11th Cir. 1984) (finding no ineffective assistance where a defense proposed on collateral review would have “conflicted” with the strategy employed at trial).

As this record makes abundantly clear, the mitigating evidence that was presented at Jefferson’s sentencing trial was brief and weak. Aside from a passing remark from Jefferson’s mother about his head injury, no evidence was presented to the jury about mental impairment. See also Stephens v. Kemp, 846 F.2d 642, 653-54 (11th Cir. 1988) (finding prejudice, where all the sentencing jury heard about the defendant’s mental problems were a few comments from his mother about his occasionally bizarre behavior; counsel’s failure to more fully develop the evidence on mental health “undoubtedly diminished” the impact the mother’s stray comments had on the jury).

The new mitigation evidence developed collaterally would have been largely consistent with the mitigation evidence arguments actually presented. Defense counsel’s closing argument focused on painting a sympathetic portrait of Jefferson, pleading for mercy, and suggesting that Jefferson’s hard upbringing meant that he was less responsible for his actions. Substantial evidence of brain damage surely would have been compatible with this defense. See Ferrell, 640 F.3d at 1235 (finding prejudice, in part, “because defense counsel did not pursue much of a residual doubt defense, instead mostly focusing on mercy, [meaning that]

75

mitigating evidence of Ferrell’s mental health and childhood would have easily and directly supported the approach counsel offered at sentencing”).

It is of course true, as this Court recognized in Jefferson’s previous appeal, that presenting extensive evidence regarding Jefferson’s brain damage would have been in some tension with Jefferson’s claims to innocence both at trial and during sentencing than his more generalized pleas for mercy were. See Jefferson v. Hall, 570 F.3d at 1307. However, the dangers posed by any potential inconsistency of presenting this evidence at the sentencing phase after trial counsel had argued innocence at the guilt phase seem to be relatively insignificant in this case. In Jones v. Dugger, 928 F.2d 1020 (11th Cir. 1991), this Court found no ineffective assistance of counsel because the presentation of mitigating evidence would have created inconsistencies with other elements of trial counsel’s defense strategy, and the “small benefit” of presenting the mitigating evidence was far outweighed by the risk of inconsistency. Id. at 1028. Here, in sharp contrast, the benefit of presenting the new and powerful forensic evidence would have been anything but small -- as we’ve detailed, evidence of Jefferson’s brain damage would have dramatically altered Jefferson’s sentencing profile and given the jury a compelling reason not to impose the death penalty.

Nor can we say that testimony from the State’s expert, Dr. Macciocchi, would have measurably damaged Jefferson’s case. Indeed, Macciocchi had no

76

medical expertise in neurology, was not a licensed physician, and, as Jefferson’s experts pointed out, he used improper norms in scoring Jefferson’s test results. Some of what he relied on would not even have been available at the time of Jefferson’s sentencing. And, according to Drs. Merikangas and Price, Macciocchi’s report did not in any way cause them to doubt their conclusions.

The long and short of it is that there is a reasonable probability that, had Jefferson’s counsel conducted a proper investigation and presented to the jury the extensive evidence of organic brain damage that such an investigation would have uncovered, at least one juror who would have reached a different judgment. The fundamental transformation of Jefferson’s sentencing profile catalyzed by the mitigation evidence developed on collateral review; that such evidence substantially undermined the aggravating factors relied on by the jury; and that the new evidence does not conflict with the weak mitigation evidence actually presented all lead us toward this conclusion: Jefferson has established that his Sixth Amendment right to the effective assistance of counsel was violated, and he is entitled to a new sentencing proceeding.

AFFIRMED.