Marcus Bernard Williams v. State of Alabama

73 F.4th 900

• Court: Court of Appeals for the Eleventh Circuit
• Decided: July 11, 2023
• Docket: 21-13734
• Status: Published

Opinion

USCA11 Case: 21-13734 Document: 39-1 Date Filed: 07/11/2023 Page: 1 of 32

[PUBLISH]

In the United States Court of Appeals For the Eleventh Circuit

____________________

No. 21-13734 ____________________

MARCUS BERNARD WILLIAMS, Petitioner-Appellee, versus STATE OF ALABAMA,

Respondent-Appellant.

Appeal from the United States District Court for the Northern District of Alabama D.C. Docket No. 1:07-cv-01276-KOB ____________________ USCA11 Case: 21-13734 Document: 39-1 Date Filed: 07/11/2023 Page: 2 of 32

2 Opinion of the Court 21-13734

Before WILSON, GRANT, and LAGOA, Circuit Judges. WILSON, Circuit Judge: In 1996, Marcus Bernard Williams was convicted of capital murder by an Alabama jury. The jury recommended death by ex- ecution, and the trial judge imposed the death penalty. Williams filed a petition for habeas corpus relief in the Northern District of Alabama, alleging—as relevant to this appeal—that trial counsel was ineffective during the penalty phase of his trial for failing to investigate and present mitigating evidence. The district court initially denied habeas relief on all claims, and Williams appealed. We vacated the district court’s order and remanded to the district court to determine whether Williams was entitled to an evidentiary hearing and to reconsider his failure-to- investigate claims de novo. Williams v. Alabama, 791 F.3d 1267, 1277 (11th Cir. 2015). After conducting an evidentiary hearing, the district court granted habeas relief. The State of Alabama (State) now appeals. After careful review of the record and with the ben- efit of oral argument, we affirm. I. BACKGROUND The only question presented in this appeal is whether Wil- liams’ trial attorneys were constitutionally ineffective during the penalty phase of his trial. Neither the facts of Williams’ case nor his guilt is in dispute. The relevant facts are as follows: On November 6th, 1996, the defendant had been out with friends, drinking and smoking marijuana. Upon returning home that evening, the defendant's USCA11 Case: 21-13734 Document: 39-1 Date Filed: 07/11/2023 Page: 3 of 32

21-13734 Opinion of the Court 3

thoughts turned to a young female neighbor of his, Melanie Dawn Rowell, and his desire to have sexual relations with her.

At approximately 1:00 a.m. that night, Williams at- tempted to enter Rowell’s back door, but the door was locked. He then noticed a kitchen window beside the door. He removed the screen from the window and found that the window was not locked.

Williams v. State, 795 So. 2d 753, 761 (Ala. Crim. App. 1999) (Wil- liams I). Williams entered through the kitchen window, took a knife from the counter, and proceeded upstairs, removing his pants part way up the stairs. Id. Williams entered Rowell’s bedroom. Id. He climbed in bed on top of her. When he began re- moving Rowell’s clothes, a struggle ensued. . . . As Rowell continued to struggle, Williams placed his hands around her neck. Eventually Rowell ceased to struggle as Williams continued to strangle her. When she was motionless, Williams proceeded to have sex- ual intercourse with her for 15 to 20 minutes.

Id. at 761–62. Williams was subsequently arrested. Id. at 762. He provided written statements to the police implicating himself in Rowell’s death. Williams was indicted for murder during a rape or attempted rape. See Ala. Code § 13A-5-40(a)(3) (1975). Trial Williams was represented at trial by Tommie Wilson (now deceased) and Erskine Funderburg. Wilson was primarily USCA11 Case: 21-13734 Document: 39-1 Date Filed: 07/11/2023 Page: 4 of 32

4 Opinion of the Court 21-13734

responsible for the guilt phase, while Funderburg mainly handled the penalty phase. After his first trial ended in a mistrial, Williams’ retrial began on February 22, 1999. Williams’ sole defense was that he intended only to rape, not kill, Rowell. The defense offered no evidence or witnesses at the guilt phase. The jury found Williams guilty of capital murder during a rape. Williams I, 795 So. 2d at 761. At the penalty phase, trial counsel called two witnesses: Wil- liams’ mother, Charlene Williams, and his great-aunt, Eloise Wil- liams. Charlene1 told the jury that she was young and unmarried when she had Williams; that Williams faced hardships as a child, such as living with different relatives and having no father or adult male figure in his life; and that Williams stopped playing school sports after a knee injury. Charlene also gave testimony that por- trayed Williams in a negative light. She testified that Williams dropped out of high school, hung out “with a rough crowd,” was kicked out of the Job Corps for fighting, stopped going to church, and “wanted to sleep all day and stay up all night.” Eloise told the jury about Williams’ unstable home life and testified that he did not see his mother often and became sad and withdrawn at times, that he was a good student with no significant criminal history, that he struggled following the deaths of his grandfather and uncle, and that since going to jail, he stayed out of trouble and was remorseful for his crime. Eloise also testified that

1 Williams’ relatives are referred to throughout this opinion by their first names to avoid confusion. USCA11 Case: 21-13734 Document: 39-1 Date Filed: 07/11/2023 Page: 5 of 32

21-13734 Opinion of the Court 5

Williams had a quick temper, had a prior arrest for fighting as a teenager, was irregularly employed after high school, and started drinking and using drugs not long before the crime. After 30 minutes of deliberation, the jury recommended a death sentence by an 11-to-1 vote. Id. At the sentencing hearing, Williams took the stand to testify, and he expressed his remorse. The trial court found one aggravat- ing circumstance—the murder was committed during a rape—and one statutory mitigating circumstance—Williams had no signifi- cant history of prior criminal activity. Id. at 784. The trial court also found four non-statutory mitigating circumstances: (1) Wil- liams’ unstable upbringing, (2) his problem resulting from the end of a promising athletic career, (3) the attainment of his GED after not graduating from high school, and (4) his remorse. Id. at 784– 85. The trial court found the aggravating factor outweighed the mitigating factors and sentenced Williams to death. Id. at 761. Direct Appeal Williams appealed to the Alabama Court of Criminal Ap- peals (ACCA). He raised several issues, including ineffective assis- tance of counsel during trial. Id. at 782. The ACCA reviewed Wil- liams’ claims for plain error because he “did not first present [them] to the trial court in a motion for a new trial.” Id. The ACCA con- cluded that Williams had not shown that his trial counsel’s perfor- mance was deficient and that their allegedly deficient performance prejudiced him. Id. at 784. The Alabama Supreme Court affirmed the ACCA’s judgment and Williams’ conviction and sentence. Ex USCA11 Case: 21-13734 Document: 39-1 Date Filed: 07/11/2023 Page: 6 of 32

6 Opinion of the Court 21-13734

parte Williams, 795 So. 2d. 785, 787–88 (Ala. 2001). The United States Supreme Court denied certiorari. Williams v. Alabama, 534 U.S. 900 (2001). State Postconviction Proceedings Williams next sought state postconviction relief based on his claim that trial counsel failed to conduct an adequate investigation into his background, thus failing to uncover a history of neglect and childhood sexual abuse by an older boy. Williams requested dis- covery and an evidentiary hearing. The postconviction court de- nied relief. The ACCA affirmed, see Williams v. State, 2 So. 3d 934 (Ala. Crim. App. 2006) (table), and the Alabama Supreme Court de- nied certiorari, see Ex parte Williams, 13 So. 3d 52 (Ala. 2007) (table). Federal Habeas Proceedings In July 2007, Williams petitioned for federal habeas corpus relief under 28 U.S.C. § 2254 from the Northern District of Ala- bama. The district court, applying deference under the Antiterror- ism and Effective Death Penalty Act (AEDPA) to the state postcon- viction court’s ruling, concluded that the ruling was not contrary to, or an unreasonable application of Strickland v. Washington, 466 U.S. 668 (1984), and denied his petition. See Williams v. Alabama, No. 1:07-cv-1276, 2012 WL 1339905 (N.D. Ala. Apr. 12, 2012). After the district court denied Williams’ request for a certificate of ap- pealability (COA), Williams filed a motion in our court seeking a COA. We granted a limited COA on Williams’ claim of ineffective assistance of counsel for failure to investigate and present USCA11 Case: 21-13734 Document: 39-1 Date Filed: 07/11/2023 Page: 7 of 32

21-13734 Opinion of the Court 7

mitigating evidence during the penalty phase of his trial. After con- sidering Williams’ appeal, we concluded that “Williams’ failure-to- investigate claims were fairly presented in state court, [but] they were not decided ‘on the merits’ within the meaning of 28 U.S.C. § 2254(d).” Williams v. Alabama, 791 F.3d 1267, 1269 (11th Cir. 2015) (Williams II). We held that the district court erred in granting deference under AEDPA to the postconviction court’s decision be- cause the ACCA had applied a procedural bar and therefore had not adjudicated Williams’ claims on the merits. Id. at 1273–74. Thus, we vacated the district court’s order denying the failure-to- investigate claims and Williams’ request for an evidentiary hearing. Id. at 1277. We remanded Williams’ case back to the district court to determine whether Williams was entitled to an evidentiary hear- ing and to reconsider his failure-to-investigate claims de novo. Id. Evidentiary Hearing On remand, the district court conducted a three-day eviden- tiary hearing on Williams’ failure-to-investigate claims.2 Williams testified at the hearing and called several witnesses who testified to

2 Williams asserted eight failure-to-investigate claims in his amended habeas petition: (1) failure to collect documentary evidence and hire a mitigation spe- cialist; (2) failure to thoroughly investigate Williams’ history, including his childhood sexual abuse; (3) failure to interview Williams’ friend, Alister Cook; (4) failure to adequately interview and prepare the penalty phase witnesses; (5) failure to compile Williams’ history of abuse and neglect; (6) failure to investi- gate Williams’ family history of mental illness; (7) failure to show that Wil- liams’ background contributed to his committing capital murder; and (8) fail- ure to present his redeeming characteristics. USCA11 Case: 21-13734 Document: 39-1 Date Filed: 07/11/2023 Page: 8 of 32

8 Opinion of the Court 21-13734

facts that Williams claims trial counsel should have presented dur- ing the penalty phase of his trial. The following lay witnesses tes- tified: attorney Funderburg; Tina Watson (attorney Wilson’s for- mer legal secretary); Billy Stephens (a mitigation investigator whom Wilson purportedly sought to hire); Sharenda and LaCharo Williams (Williams’ sisters); Marlon Bothwell (Williams’ child- hood friend); Eloise; and Charlene. Williams also presented testi- mony from clinical psychologist Dr. Matthew Mendel and neuro- psychologist Dr. Kenneth Benedict to explain the effects that child sexual abuse, alcoholism, abandonment, and familial dysfunction had on Williams. Wilson did not testify because she passed away in 2015. However, Wilson’s case file was presented as documen- tary evidence. The State presented expert testimony from Dr. Glen King to counter Williams’ experts. Both sides presented eviden- tiary materials. Williams argued that, had counsel performed an adequate penalty phase mitigation investigation, they would have discov- ered and been able to present evidence regarding: childhood sexual abuse by an older boy; an extensive family history of childhood sex- ual abuse and incest; a family history of alcoholism which contrib- uted to Williams’ early and excessive use of alcohol; a childhood defined by chaos, abandonment, and abuse; an extensive family his- tory of fracture and dysfunction; and psychologically damaging ex- periences during childhood. On September 23, 2021, the district court entered a 141-page order granting Williams’ habeas petition on all his failure-to- USCA11 Case: 21-13734 Document: 39-1 Date Filed: 07/11/2023 Page: 9 of 32

21-13734 Opinion of the Court 9

investigate claims except for three of them.3 Williams v. Alabama, No. 1:07-cv-1276, 2021 WL 4325693, at *1 (N.D. Ala. Sept. 23, 2021) (Williams III). The State timely appealed. II. STANDARD OF REVIEW We review de novo the district court’s grant of a federal ha- beas petition under 28 U.S.C. § 2254. Peterka v. McNeil, 532 F.3d 1199, 1200 (11th Cir. 2008). The district court’s factual findings in a habeas proceeding are reviewed for clear error—a highly defer- ential standard of review. Sims v. Singletary, 155 F.3d 1297, 1304 (11th Cir. 1998); Holton v. City of Thomasville Sch. Dist., 425 F.3d 1325, 1350 (11th Cir. 2005). “A finding is ‘clearly erroneous’ when although there is evidence to support it, the reviewing court on the entire evidence is left with the definite and firm conviction that a mistake has been committed.” Anderson v. City of Bessemer City, 470 U.S. 564, 573 (1985) (alteration adopted) (quoting United States v. United States Gypsum Co., 333 U.S. 364, 395 (1948)). “An ineffective assistance of counsel claim is a mixed question of law and fact which we review de novo.” Sims, 155 F.3d at 1304.

3 The claims that the district court did not grant habeas relief on were “coun- sels’ failure to interview Mr. Williams’ closest friend Alister Cook, failure to investigate his family history of mental illness, and the failure to present his redeeming characteristics because Mr. Williams . . . failed to show prejudice on these three claims.” Williams III, 2021 WL 4325693, at *1. USCA11 Case: 21-13734 Document: 39-1 Date Filed: 07/11/2023 Page: 10 of 32

10 Opinion of the Court 21-13734

III. DISCUSSION The State argues on appeal that the district court erred in two ways. First, by failing to uphold the presumption of effective assistance of counsel, and second, by incorrectly reweighing the ad- ditional aggravating and mitigating evidence produced at the evi- dentiary hearing. None of the State’s arguments has merit. To succeed on an ineffective assistance of counsel claim, a defendant must demonstrate two things: (1) that counsel’s perfor- mance was deficient, and (2) that the deficiency prejudiced the de- fense. Strickland, 466 U.S. at 687. A. Deficient Performance To establish deficient performance, a defendant must show “that counsel made errors so serious that counsel was not function- ing as the ‘counsel’ guaranteed the defendant by the Sixth Amend- ment.” Id. “[T]he proper standard for attorney performance is that of reasonably effective assistance.” Id. Thus, to prevail on an inef- fectiveness claim, “the defendant must show that counsel’s repre- sentation fell below an objective standard of reasonableness.” Id. at 688. A petitioner bears the burden of proving counsel’s perfor- mance was unreasonable by “a preponderance of competent evi- dence.” Chandler v. United States, 218 F.3d 1305, 1313 (11th Cir. 2000) (en banc). Regarding the duty to investigate, “counsel has a duty to make a reasonable investigation or to make a reasonable decision that makes particular investigations unnecessary. . . . [A] particular decision not to investigate must be directly assessed for USCA11 Case: 21-13734 Document: 39-1 Date Filed: 07/11/2023 Page: 11 of 32

21-13734 Opinion of the Court 11

reasonableness in all the circumstances, applying a heavy measure of deference to counsel’s judgments.” Strickland, 466 U.S. at 691. The district court found “that counsels’ performance was deficient for failing to reasonably investigate Mr. Williams’ back- ground for mitigation.” Williams III, 2021 WL 4325693, at *17. In making this determination, the district court considered “what Mr. Williams’ counsel knew about him, his criminal charges, and his background and ‘what counsel then failed to do and learn about [Mr. Williams] and his childhood background.’” Id. (citing Hard- wick v. Sec’y, Fla. Dep’t of Corr., 803 F.3d 541, 552 (11th Cir. 2015). Because the State has not shown any of the district court’s factual findings are clearly erroneous, we agree with the district court. After reviewing the totality of the evidence in the record and produced at the evidentiary hearing, we find that Funderburg’s and Wilson’s representation of Williams at the penalty phase “fell be- low an objective standard of reasonableness.” Strickland, 466 U.S. at 687–88. Specifically, counsel failed to conduct an adequate in- vestigation into Williams’ background for possible mitigating evi- dence. Counsel knew how important the penalty phase of the trial was, given the overwhelming evidence of Williams’ guilt. Funder- burg testified that he knew Williams’ confession would likely be admitted at trial, so mitigation would be important to possibly get life without parole. Wilson also knew the importance of mitiga- tion. Handwritten notes from her case file show that she identified the “best issue” to be whether the sentence would be life without parole or death. Counsel also knew the sexual nature of the crime USCA11 Case: 21-13734 Document: 39-1 Date Filed: 07/11/2023 Page: 12 of 32

12 Opinion of the Court 21-13734

and that alcohol and marijuana played a significant role, so reason- able counsel with this knowledge would have thoroughly investi- gated Williams’ sexual history and how he began abusing alcohol and marijuana. Despite counsel’s knowledge that the penalty phase would be crucial, counsel failed to use available resources for a mitigation investigation. The trial court had granted a motion from Wilson and awarded $1,500 to hire a mitigation investigator for the penalty phase, but counsel never used the court-awarded funds to retain a mitigation investigator. The evidence also shows that counsel unreasonably delayed starting their mitigation investigation, and when counsel did start at the eleventh hour, their efforts were minimal and deficient. The trial court appointed Wilson in November 1996 and Funderburg in May 1997, and Williams’ trial was ultimately set for November 1998. Yet, at the time Wilson filed the June 1997 motion seeking funds for a mitigation investigator—seven months after her ap- pointment—she stated in the motion that there had “not been ad- equate investigation into critical matters relevant to . . . [the] level of culpability and appropriate punishment.” By August 1997, Fun- derburg had not spoken to any of Williams’ friends or family re- garding his background. And by February 1999—only days prior to Williams’ trial—the only documentary records that Funderburg had received and reviewed were Williams’ Job Corps records. Funderburg testified at the evidentiary hearing that he did not contact any family or friends prior to August 1997 because he USCA11 Case: 21-13734 Document: 39-1 Date Filed: 07/11/2023 Page: 13 of 32

21-13734 Opinion of the Court 13

did not think there was much cooperation at the time from Wil- liams’ friends who were also suspects. But counsel did not even reach out to those friends or relatives who were not suspects, such as Williams’ childhood friend, Marlon Bothwell, or his sisters, LaCharo and Sharenda. Prior to trial, the only relatives Funder- burg met with were Williams’ mother, Charlene, and his great- grandmother, Beulah Williams. Counsel called only two witnesses at the penalty phase—Charlene and Eloise. Funderburg failed to properly prepare Eloise to testify, as he only met with her for the first time on the day of her testimony for about fifteen minutes. Eloise testified at the evidentiary hearing that counsel did not seem to understand Williams’ life story. The evidence shows that counsel met infrequently with Wil- liams and failed to ask more than general questions about Wil- liams’ background. Williams testified at the evidentiary hearing that he met with counsel only about “a half dozen times,” and the meetings lasted 15 to 30 minutes. Counsel’s fee declarations sup- port Williams’ testimony. Wilson’s fee declaration reflects that she met with Williams only twice for a total of 3 hours over the course of two years. Funderburg’s fee declaration shows he met with Wil- liams five times for a total of 8 hours (although 4.5 of the 8 hours were conferences with Williams and the District Attorney, so pre- sumably Funderburg and Williams were not alone). Williams tes- tified that Funderburg asked him only general questions about his childhood, such as about family and school. Williams was never asked about his family background; whether he had been USCA11 Case: 21-13734 Document: 39-1 Date Filed: 07/11/2023 Page: 14 of 32

14 Opinion of the Court 21-13734

neglected; or whether he had been sexually, physically, or emo- tionally abused. These deficiencies were patently unreasonable. See Johnson v. Sec’y, DOC, 643 F.3d 907, 932 (11th Cir. 2011) (“Given the over- whelming evidence of guilt, any reasonable attorney would have known . . . that the sentence stage was the only part of the trial in which [the defendant] had any reasonable chance of success.”) Had counsel conducted a more thorough investigation into Williams’ sexual history, they would have learned that Williams had been sexually abused on three or four occasions between the ages of four and six when he and his mother lived with the Mostella family. At the evidentiary hearing, Williams testified that Mario Mostella, who was older 4 and used to babysit him, used to pretend they were playing a game in which they would touch each other’s genitals, and Mostella would anally penetrate Williams. Williams never told his mother about the sexual abuse, and the abuse led Williams to have thoughts of self-harm and suicide and feelings of shame and depression. Williams stated that he never told his trial attorneys about the sexual abuse because “[t]hey didn’t ask,” but he later told his postconviction attorney. Williams testified that he did not tell his postconviction attorney about the sexual abuse

4 Williams testified that he believed Mostella was ten or twelve years older than him, whereas Eloise testified that Mostella was “maybe ten” years old. While Mostella’s exact age is unclear, it is apparent that he was older than Williams and, since he used to babysit Williams, presumably would have been in control when they were unsupervised. USCA11 Case: 21-13734 Document: 39-1 Date Filed: 07/11/2023 Page: 15 of 32

21-13734 Opinion of the Court 15

when first asked, but after growing comfortable with him over time, Williams shared that he had been sexually abused when counsel later asked him again. Williams was also exposed to sexual relations at an inappro- priately young age. There were poor boundaries in Williams’ fam- ily with regard to sexuality. For example, Charlene used to have her boyfriends in the same bed that she shared with Williams (alt- hough Williams never witnessed his mother having sexual rela- tions). Williams testified that when he was ten years old, his 18- or 19-year-old cousin Brian Williams allowed him “to watch [Brian] have sex as a way of showing [Williams] how to do it with a woman.” Williams’ sexual abuse and his exposure to other peo- ple’s sexual relations at such a young age had a significant detri- mental impact on his own sexual development. Dr. Mendel testi- fied that male victims of sexual abuse will often display “compul- sive sexuality or hyper sexuality, [become] very driven to be sex- ually active and have numerous partners.” Williams became sex- ually active at the age of ten, had about 75 sexual partners by the time he graduated high school, and about 150 sexual partners by the time of his arrest. Williams had no serious or committed rela- tionships, just a “pattern of repeated hookups and sexual encoun- ters.” Dr. Mendel testified that Williams’ promiscuity was “very reassuring to him” and “made him feel like a man.” Had counsel conducted an adequate investigation into Wil- liams’ family background, they also would have learned about the domestic violence Williams witnessed between his mother and her USCA11 Case: 21-13734 Document: 39-1 Date Filed: 07/11/2023 Page: 16 of 32

16 Opinion of the Court 21-13734

abusive boyfriend, Jeff Deavers. Williams saw his mother “with a black eye and busted lip,” and on another occasion he saw Deavers strike his mother with his hand. Williams, who was then 12 or 13 years old, grabbed a knife and tried to stab Deavers. Counsel would have also learned about the pervasive his- tory of childhood sexual abuse and incest within Williams’ family, which spans multiple generations. Dr. Mendel stated in his expert report that Williams’ great-grandmother Beulah was raped by her uncle, who fathered her child; his grandmother Laura’s first child was fathered by Laura’s cousin; his aunt Veronica was molested by her aunt Helen’s boyfriend; and his cousin Brian molested Wil- liams’ sister LaCharo and his cousin Zakia. Within Williams’ fam- ily, the sexual abuse and molestations were simply “swept under the rug.” According to Dr. Mendel, within Williams’ family, “dis- closure would have been worthless, resulting in neither protection from further abuse or treatment for the impact of the abuse.” Further, counsel would have learned about the alcoholism that was rampant in Williams’ family, and how Williams’ mother Charlene often drank to the point of intoxication and neglected Williams as a child. Williams began drinking alcohol between the ages of 12 and 14 years old, and his consumption steadily increased through his teenage years to the point where he was getting drunk weekly. Finally, counsel would have learned that Williams’ child- hood was defined by chaos and abandonment. As a child, Williams and his teenage mother lacked a stable home life, instead bouncing USCA11 Case: 21-13734 Document: 39-1 Date Filed: 07/11/2023 Page: 17 of 32

21-13734 Opinion of the Court 17

around to live with different relatives or family friends. Eloise tes- tified at the evidentiary hearing that Charlene would often go out partying or drinking; she would leave the children to fend for them- selves; and at times the children were not clean or well cared for. When Williams was six or seven years old, he moved back and forth between Eloise and his great-grandmother because Charlene could not properly care for him. Charlene and Williams’ father were not in a committed relationship, and Williams’ father was not involved in his life until he was 13 or 14 years old. Williams did not grow up with any of his six half-siblings, all of whom were raised in different households. Williams felt that all of his half-siblings had a place they could call home, whereas Williams lacked a consistent home and felt that he was never wanted or belonged anywhere. Had trial counsel conducted an adequate investigation into Williams’ background, they would have learned about all the cir- cumstances discussed above and been able to present them to the jury as mitigating evidence. But the jury never heard any of this compelling mitigating evidence. Because the Supreme Court has endorsed the use of the American Bar Association Guidelines for the Appointment and Per- formance of Counsel in Death Penalty Cases 1989 (ABA Death Pen- alty Guidelines), the district court properly used it to determine whether Williams’ counsel was deficient. See Wiggins v. Smith, 539 U.S. 510, 524 (2003). Looking to the prevailing norms at the time of Williams’ trial, a capital defendant who does not have a “credible argument for innocence . . . has the right to present his or her USCA11 Case: 21-13734 Document: 39-1 Date Filed: 07/11/2023 Page: 18 of 32

18 Opinion of the Court 21-13734

sentencer with any mitigating evidence that might save his or her life.” ABA Death Penalty Guidelines 1.1, commentary. The ABA Death Penalty Guidelines further instruct that counsel should in- vestigate both the guilt and penalty phases “immediately upon counsel’s entry into the case and [the investigation] should be pur- sued expeditiously.” Id. 11.4.1(A). Moreover, as soon as appropri- ate counsel should “collect information relevant to the sentencing phase of trial including, but not limited to . . . family and social history (including physical, sexual or emotional abuse).” Id. 11.4.1(2)(C). It is clear to us from the totality of the evidence that counsel spent minimal time and effort conducting a background investiga- tion for potential mitigating evidence that would help the jury un- derstand why Williams committed the crime that he did. Coun- sel’s minimal efforts were also unreasonably delayed and untimely. As a result, counsel’s failure to conduct an adequate background investigation for mitigating evidence deprived Williams of reason- ably effective assistance. Strickland, 466 U.S. at 687. The State argues that the district court failed to uphold the presumption of effective assistance when it found that attorney Wilson unreasonably failed to keep better records. Citing Callahan v. Campbell, 427 F.3d 897, 933 (11th Cir. 2005), the State asserts that when assessing a deceased attorney’s performance, courts presume the attorney “did what [s]he should have done” and exercised rea- sonable professional judgment. We are not persuaded by the State’s argument. While it is true that “[j]udicial scrutiny of USCA11 Case: 21-13734 Document: 39-1 Date Filed: 07/11/2023 Page: 19 of 32

21-13734 Opinion of the Court 19

counsel’s performance must be highly deferential” and we must endeavor “to eliminate the distorting effects of hindsight,” Strick- land, 466 U.S. at 680, the deferential standard is not insurmounta- ble. Although Wilson could not testify herself, we have a thorough understanding from the other witnesses and documentary evi- dence of what Wilson did—and, more importantly, did not do—as far as Williams’ mitigation defense during the penalty phase. Fun- derburg, Wilson’s co-counsel, and Tina Watson, her legal assistant of almost two decades, both testified at the evidentiary hearing, and Wilson’s case file was also produced. Here, there is abundant evidence that trial counsel’s performance at the penalty phase was untimely, deficient, and unreasonable. Nor were counsel’s omissions the result of strategy. Trial counsel had “a duty to make reasonable investigations or to make a reasonable decision that makes particular investigations unneces- sary.” Strickland, 466 U.S. at 691. Here, counsel did not conduct a reasonable background investigation, and the decision not to inves- tigate was not the result of any strategic choice. In order to make a strategic choice about which evidence to present and which evi- dence to omit, counsel needed to first investigate and discover the evidence and then make an informed, strategic decision. Here, counsel simply failed to conduct an adequate investigation in the first place. In sum, the district court made thorough factual findings af- ter the evidentiary hearing, which help us assess the merit of Wil- liams’ failure-to-investigate claims—and the State has not USCA11 Case: 21-13734 Document: 39-1 Date Filed: 07/11/2023 Page: 20 of 32

20 Opinion of the Court 21-13734

specifically identified any of those factual findings as clearly errone- ous. Considering the district court’s factual findings and the total- ity of the evidence in the record, on de novo review we find that “in light of all the circumstances,” trial counsel’s failure to investi- gate Williams’ background for mitigating evidence was “outside the wide range of professionally competent assistance.” Id. at 690. B. Prejudice In addition to deficient performance, a petitioner must also establish prejudice to succeed on an ineffective-assistance-of-coun- sel claim. Id. at 687. To establish prejudice, a defendant must show “that counsel’s errors were so serious as to deprive the defendant of a fair trial, a trial whose result is reliable.” Id. When a defendant challenges a death sentence, the test for prejudice “is whether there is a reasonable probability that, absent [counsel’s] errors, the sen- tencer—including an appellate court, to the extent it independently reweighs the evidence—would have concluded that the balance of aggravating and mitigating circumstances did not warrant death.” Id. at 695. “A reasonable probability is a probability sufficient to undermine confidence in the outcome.” Id. at 694. After carefully reweighing the evidence, we find there is a reasonable probability that, absent counsel’s deficiencies, the bal- ance of aggravating and mitigating factors in Williams’ case did not warrant a sentence of death. At the penalty phase of Williams’ trial, the jury only heard testimony from Eloise and Charlene. However, their testimony revealed very little about the true extent of Williams’ troubled USCA11 Case: 21-13734 Document: 39-1 Date Filed: 07/11/2023 Page: 21 of 32

21-13734 Opinion of the Court 21

upbringing and family history. The jury never heard about Wil- liams’ sexual abuse, his early exposure to sexual relations, his expo- sure to domestic violence, the abandonment by both his father and mother, or the sexual abuse and alcoholism that was pervasive in his family. The Supreme Court has found that counsel’s failure to present evidence of abuse in mitigation constitutes prejudice. See, e.g., Wiggins, 539 U.S. at 534–35 (finding prejudice where counsel failed to discover and present mitigating evidence of the defend- ant’s physical and sexual abuse during childhood); Rompilla v. Beard, 545 U.S. 374, 391–93 (2005) (finding petitioner was prejudiced by counsel’s failure to present evidence of parents’ alcoholism, domes- tic violence, physical and verbal abuse, and dire living conditions). In Williams’ case, “the nature, quality, and volume of the mitiga- tion never known to the jury is significant enough to conclude that it ‘bears no relation’ to the cursory evidence that trial counsel pre- sented.” Daniel v. Comm’r, Ala. Dep’t of Corr., 822 F.3d 1248, 1276 (11th Cir. 2016) (quoting Rompilla, 545 U.S. at 393). Had the jury been presented with all of the mitigating evidence, there is a rea- sonable probability that the outcome of Williams’ trial could have been different. Strickland, 466 U.S. at 695. The State argues that the district court incorrectly re- weighed the additional aggravating and mitigating evidence pro- duced at the evidentiary hearing by giving significance to the num- ber of aggravators and mitigators rather than their nature. This argument is meritless and, in all events, on de novo review, this Court has independently reweighed all the available evidence. See Strickland, 466 U.S. at 695. We have reviewed the record and the USCA11 Case: 21-13734 Document: 39-1 Date Filed: 07/11/2023 Page: 22 of 32

22 Opinion of the Court 21-13734

evidence produced at the evidentiary hearing, and we find that not only the sheer volume of but also the powerful nature of the miti- gators overwhelmingly outweighs the aggravator in Williams’ case. We do not minimize the weight or significance of the aggra- vating circumstance—that the murder was committed during a rape—but, when balancing it against all of the mitigating circum- stances that we now know, our confidence in the outcome of the penalty phase of Williams’ trial is undermined. Id. at 694. The State also asserts that the district court erroneously found Williams was prejudiced by counsel’s failure to present evi- dence of childhood sexual abuse by Mostella because Williams never described the experience as traumatic, and the mitigation value of the evidence is weakened by the passage of time between the abuse and Williams’ crime. We disagree. Dr. Mendel testified at the evidentiary hearing that he strongly believed that “if the sex- ual abuse hadn’t happened, there would not have been the sexual violence.” The abuse clearly had a damaging impact on Williams because he felt shameful and had thoughts of hurting or killing himself. The abuse also contributed to the early age at which Wil- liams became sexually active, and the hypersexuality he developed as an adolescent and a young man. The district court credited both Williams’ and Dr. Mendel’s testimony regarding the sexual abuse by Mostella, and the State has not adequately challenged this cred- ibility finding. Thus, we reject the State’s argument that Williams was not prejudiced by the failure to present evidence regarding Williams’ sexual abuse by Mostella. See Williams II, 791 F.3d at 1277 (noting that sexual molestation and repeated rape “during USCA11 Case: 21-13734 Document: 39-1 Date Filed: 07/11/2023 Page: 23 of 32

21-13734 Opinion of the Court 23

childhood can be powerful mitigating evidence, and is precisely the type of evidence that is ‘relevant to assessing a defendant’s moral culpability’”). The dissent disagrees that Williams has established preju- dice and places much emphasis on the facts of the underlying mur- der. Dissenting Op. at 2–3. While we do not minimize the brutal- ity of Williams’ crime, those facts must be weighed against all the mitigating evidence. Here, the district court conducted an eviden- tiary hearing on the failure-to-investigate claims, made extensive factual findings based on evidence that had not been presented dur- ing Williams’ penalty phase, and concluded that Williams was en- titled to habeas relief. On appeal, we must review the district court’s factual findings for clear error and its legal conclusions de novo. The clear error standard is highly deferential. Holton, 425 F.3d at 1350. The dissent has not suggested that any of the district court’s factual findings were clearly erroneous. We find no clear error in the court’s factual findings. Considering the record before us—and given the highly deferential standard of review for factual findings—we conclude that Williams has established Strickland prejudice. Thus, Williams “has met the burden of showing that the de- cision reached [at the penalty phase] would reasonably likely have been different absent the errors.” Strickland, 466 U.S. at 696. IV. CONCLUSION Williams has established both that his trial counsel rendered deficient performance during the penalty phase of his trial, and that USCA11 Case: 21-13734 Document: 39-1 Date Filed: 07/11/2023 Page: 24 of 32

24 Opinion of the Court 21-13734

their deficient performance prejudiced him. As a result, Williams’ trial was fundamentally unfair. Accordingly, we affirm the district court’s grant of habeas corpus relief. AFFIRMED. USCA11 Case: 21-13734 Document: 39-1 Date Filed: 07/11/2023 Page: 25 of 32

21-13734 GRANT, J., dissenting 1

GRANT, Circuit Judge, dissenting: Trial counsel’s efforts to investigate Williams’s background and prepare for the sentencing phase were unacceptable. Despite being fully aware that the strength of the evidence of their client’s guilt meant a thorough mitigation investigation was imperative, the two attorneys billed a combined total of less than 15 hours for sentencing-phase preparation. But even though I join the majority in concluding that counsel’s performance was deficient, I do not agree that Williams can meet his burden of showing a reasonable probability that, if not for counsel’s substandard performance, the sentencing authority “would have concluded that the balance of aggravating and mitigating circumstances did not warrant death.” Strickland v. Washington, 466 U.S. 668, 695 (1984). I therefore respectfully dissent. The only way to conclude that Williams met that burden is to dismiss the statutory aggravator—that the murder was committed during a burglary and rape, Alabama Code § 13A-5- 49(4)—without adequate consideration. We have previously explained that when a murder is “carefully planned, or accompanied by torture, rape or kidnapping,” the aggravating circumstances of the crime itself may “outweigh any prejudice caused when a lawyer fails to present mitigating evidence.” Callahan v. Campbell, 427 F.3d 897, 938 (11th Cir. 2005) (quoting Dobbs v. Turpin, 142 F.3d 1383, 1390 (11th Cir. 1998)). This is especially true where the murder is a brutal one—“or, even, a less USCA11 Case: 21-13734 Document: 39-1 Date Filed: 07/11/2023 Page: 26 of 32

2 GRANT, J., dissenting 21-13734

brutal murder for which there is strong evidence of guilt in fact.” Clisby v. Alabama, 26 F.3d 1054, 1057 (11th Cir. 1994). Here, the murder of Rowell was brutal. And it was proven by overwhelming evidence of Williams’s guilt, including blood and semen evidence and several statements by Williams describing his crimes. See Williams v. Alabama, 791 F.3d 1267, 1269 (11th Cir. 2015). The circumstances of the burglary, rape, and murder that Williams committed bear repeating. Melanie Rowell, a young mother with two small children, was fast asleep when Williams invaded her home in the middle of the night, intent on raping her. He took a knife from her kitchen and went quietly up the stairs, taking his pants off along the way. At the top of the stairs, he climbed over a baby gate and stopped to “peek[] in” at the children, who were asleep in the bedroom across the hall from Rowell’s. Rowell awoke with Williams on top of her, holding a knife to her throat and pulling off her shorts. She screamed and struggled and bit his hand, but he held her down and strangled her until she stopped moving, and then raped her—apparently not caring whether she was alive or dead at that point. Afterward, Williams left Rowell’s brutalized and half-naked body on the floor, where it was discovered first by Rowell’s 15-month-old daughter and then by her mother. Any evaluation of the aggravating circumstance must acknowledge the shock and terror of the home invasion, the added fear the victim must have felt for her children, the cold brutality shown by Williams in strangling a neighbor to death so USCA11 Case: 21-13734 Document: 39-1 Date Filed: 07/11/2023 Page: 27 of 32

21-13734 GRANT, J., dissenting 3

that he could rape her, and the cruel aftermath of the murder for Rowell’s family, especially her young children. See Miller v. State, 913 So. 2d 1148, 1164 (Ala. Crim. App. 2004) (explaining that victim impact evidence is admissible during the penalty phase and relevant to whether the death penalty should be imposed). Here, the majority concludes that the “sheer volume” and “powerful nature” of the mitigating evidence “overwhelmingly outweighs the aggravator in Williams’ case.” Maj. op. at 22. I cannot agree. To start, the volume of mitigating evidence has little bearing on its weight, especially when much of the evidence is cumulative of the testimony given at trial. See Wong v. Belmontes, 558 U.S. 15, 22–23 (2009) (finding no prejudice where some of the new evidence was cumulative of evidence presented at trial and thus of little use); Holsey v. Warden, Georgia Diagnostic Prison, 694 F.3d 1230, 1270–71 (11th Cir. 2012) (same). Williams’s mother and aunt testified during the original penalty phase about Williams’s difficult childhood, including his abandonment by his father, his lack of a stable home, and his sadness and sense of abandonment when his mother left him to be cared for by relatives. Their testimony about Williams was much the same during the federal habeas proceedings. That this narrative was already presented makes it no less tragic. But it does minimize the potential effect of the new evidence Williams offers. Williams’s new evidence of childhood abuse also fails to match the standard set by the Supreme Court in Wiggins v. Smith USCA11 Case: 21-13734 Document: 39-1 Date Filed: 07/11/2023 Page: 28 of 32

4 GRANT, J., dissenting 21-13734

and Rompilla v. Beard. 1 His testimony that he was sexually abused by an older boy on three or four occasions over a two-year period, though horrifying, bears little resemblance to the “severe privation and abuse,” “physical torment, sexual molestation, and repeated rape” demonstrated in Wiggins, or the long history of severe neglect and physical and emotional abuse described in Rompilla. See Wiggins v. Smith, 539 U.S. 510, 516–17, 535 (2003); Rompilla v. Beard, 545 U.S. 374, 390–93 (2005). Indeed, the State’s expert testified that Williams did not experience trauma from the encounters, and Williams’s own expert testified that the lack of stability in Williams’s home life had a more negative and traumatizing impact on Williams than the sexual abuse. Other new evidence has even less mitigating value. Testimony that members of Williams’s extended family also suffered childhood sexual abuse has little relevance—and thus should be accorded little weight—given the lack of evidence that Williams witnessed or even knew about the abuse. Evidence of widespread alcoholism among the adults in Williams’s life would have served only to partially explain Williams’s own alcoholism;

1 The majority accepts as a given that Williams would have disclosed the abuse

to trial counsel if they had asked, but I am not convinced. Williams never told anyone about the abuse until decades after it occurred and after he had already been sentenced to death. He denied childhood abuse during his pretrial psychological evaluation and when he was first asked about it by his postconviction counsel. Williams’s psychological expert opined that Williams would also have initially denied sexual abuse if his trial counsel had asked him about it and would only say that it was “possible” that the abuse “could have come out” with “time and development of trust.” USCA11 Case: 21-13734 Document: 39-1 Date Filed: 07/11/2023 Page: 29 of 32

21-13734 GRANT, J., dissenting 5

the witnesses gave no indication that his family members were violent or abusive when drunk. And more detailed evidence of Williams’s own drug and alcohol use likely would not have helped him at all. Such evidence “often has little mitigating value and can do as much or more harm than good in the eyes of the jury.” Crawford v. Head, 311 F.3d 1288, 1321 (11th Cir. 2002); see, e.g., Grayson v. Thompson, 257 F.3d 1194, 1227 (11th Cir. 2001). One more thing to consider is that Williams’s evidence related to childhood sexual abuse is not entirely mitigating, and would also invite unfavorable evidence in rebuttal. We “must consider all the evidence—the good and the bad—when evaluating prejudice.” Belmontes, 558 U.S. at 26. In explaining the link between the childhood sexual abuse and Williams’s capital crime, Williams’s expert testified that he compensated for feelings of shame and self-doubt by becoming hypersexual and hyperaggressive. He became sexually promiscuous without developing any stable romantic relationships. He was suspended from school twice and kicked out of Job Corps for fighting, and he was arrested for assault during his teens. Further, the new evidence shows that although Williams was open about his hypersexuality, he refused to acknowledge his “anger and the tendencies toward violence” arising from the abuse. This testimony would have had the potential to harm Williams in the eyes of the jury. It would have also invited argument by the State that Williams’s tendencies toward violence and aggression would make him a danger to other inmates if he USCA11 Case: 21-13734 Document: 39-1 Date Filed: 07/11/2023 Page: 30 of 32

6 GRANT, J., dissenting 21-13734

were released into the general population to serve a life sentence. See Simmons v. South Carolina, 512 U.S. 154, 165 n.5 (1994) (plurality opinion) (“the fact that a defendant is parole ineligible does not prevent the State from arguing that the defendant poses a future danger”); see also id. at 177 (O’Connor, J., concurring in judgment) (when the defendant raises his parole ineligibility, “the prosecution is free to argue that the defendant would be dangerous in prison”). To prove this propensity, the State undoubtedly would have introduced evidence that before his arrest, Williams committed a second burglary and sexual assault on another neighbor—this time a woman he had known since childhood. The evidence would have shown that only a few weeks after murdering Rowell, Williams again invaded a neighbor’s home in the middle of the night with the intention of raping her. The record reflects graphic evidence of yet another violent attack in which Williams took off his pants, climbed through a window, and attacked the woman in her bed, holding her down while he rubbed his penis on her and “put his hand up in her vagina” as she struggled and begged for her life. When Williams discovered that the woman was menstruating, he ordered her to perform oral sex on him. She continued to struggle with Williams until daybreak, when he finally left. This evidence is relevant in several ways. To start, it is admissible in Alabama to show future dangerousness, which is “a subject of inestimable concern at the penalty phase.” Floyd v. State, 289 So. 3d 337, 431 (Ala. Crim. App. 2017) (citation omitted). And USCA11 Case: 21-13734 Document: 39-1 Date Filed: 07/11/2023 Page: 31 of 32

21-13734 GRANT, J., dissenting 7

while future dangerousness is not an aggravating circumstance in itself, it is relevant to determining the weight that should be afforded to the aggravating circumstance proven by the State. See id. The evidence of Williams’s second burglary and attempted rape would have also reduced or eliminated the mitigating value of his lack of significant prior criminal history. Just as significantly, the later crime would have demonstrated Williams’s lack of regret in the weeks following Rowell’s murder, undermining his (already unconvincing) claim for the mitigating circumstance of remorse. In addition to the previously unexplored evidence depicting Williams as an unrepentant murderer and serial home invader, the State likely would have responded to Williams’s claims of childhood suffering by further developing evidence of the lasting trauma Williams inflicted on Rowell’s children by killing their mother. At the initial sentencing hearing, the State introduced only one witness to testify about the children’s anger, fear, grief, and confusion; the evidence Williams now proffers would have opened the door to much more. See, e.g., Woodward v. State, 123 So. 3d 989, 1041–43 (Ala. Crim. App. 2011), as modified on denial of reh’g (Aug. 24, 2012). Given all of these facts, I do not agree that it is reasonably likely that the assistance of competent counsel at trial would have resulted in a different sentence. Strickland, 466 U.S. at 696. To be sure, I do not excuse the failure of the defense team to properly investigate and present a mitigation case. But viewed in its entirety USCA11 Case: 21-13734 Document: 39-1 Date Filed: 07/11/2023 Page: 32 of 32

8 GRANT, J., dissenting 21-13734

and weighed properly, the evidence developed in habeas creates only the slightest possibility of a different outcome. This conclusion does not in any way conflict with the factual findings made by the district court. Contra Maj. op. at 23. And the district court’s final determination on the prejudice prong is a mixed question of law and fact that we review de novo. Strickland, 466 U.S. at 698; Jefferson v. GDCP Warden, 941 F.3d 452, 473 (11th Cir. 2019). To show prejudice under Strickland, the “likelihood of a different result must be substantial, not just conceivable.” Harrington v. Richter, 562 U.S. 86, 111–12 (2011). Because Williams has not met that standard, I would reverse the district court’s grant of federal habeas relief and reinstate Williams’s death sentence.