Case: 17-14570    Date Filed: 07/13/2020     Page: 1 of 56



                                                                        [PUBLISH]



             IN THE UNITED STATES COURT OF APPEALS

                     FOR THE ELEVENTH CIRCUIT
                       ________________________

                               No. 17-14570
                         ________________________

                D.C. Docket No. 2:02-cv-00777-WKW-SRW



DONALD DALLAS,

                                                 Petitioner - Appellant,

                                   versus

WARDEN,
ATTORNEY GENERAL, STATE OF ALABAMA,
COMMISSIONER, ALABAMA DEPARTMENT OF CORRECTIONS,

                                                 Respondents - Appellees.

                         ________________________

                Appeal from the United States District Court
                    for the Middle District of Alabama
                      ________________________

                               (July 13, 2020)

Before MARTIN, BRANCH and MARCUS, Circuit Judges.

MARCUS, Circuit Judge:
              Case: 17-14570     Date Filed: 07/13/2020   Page: 2 of 56



      On a hot July day in 1994, Donald Dallas and Carolyn (“Polly”) Yaw

hatched a plan to obtain money to buy crack cocaine. At a shopping center in

Prattville, Alabama, they chose their victim, 73-year-old Hazel Liveoak. As Mrs.

Liveoak loaded groceries into her car, Dallas and Yaw approached the vehicle.

They pushed her into the car and forced her to lie down on the floorboard. When

they discovered that she had only $10 in cash, they abducted her instead. They

drove her first about an hour away to the end of a dirt road in Greenville, Alabama,

where Dallas demanded Liveoak’s credit cards and put her in the trunk of the car.

Dallas and Yaw then drove back to the Am-South Bank in a K-Mart parking lot in

Montgomery, where they used her bank cards to withdraw money from an ATM.

      As Yaw withdrew the money, Dallas sat on the trunk of the car and spoke

with the victim, who told him she had a heart condition. She also told him she had

a son Dallas could call who could release her from the car. She gave Dallas the

telephone number, but he didn’t write it down. He promised Liveoak that he

would call the police to make sure she was released unharmed, but he never did.

Instead, Dallas and Yaw called a cab and went immediately to a crack house to buy

drugs, and then to a motel to smoke crack all night. Meanwhile, Hazel Liveoak

struggled for hours to get free or call for help. She eventually succumbed to a

heart attack in the hot trunk of her own car.




                                          2
              Case: 17-14570     Date Filed: 07/13/2020    Page: 3 of 56



      An Alabama jury convicted Dallas of capital murder, concluding that he

intended for Hazel Liveoak to die in that trunk. The trial court accepted the jury’s

11 to 1 recommendation for death and sentenced Dallas to die for the murder. His

convictions and sentence were affirmed on direct appeal, his state postconviction

petition was denied, and the United States District Court for the Middle District of

Alabama denied his federal habeas petition. We granted a certificate of

appealability limited to two issues: (1) whether Dallas received ineffective

assistance of counsel throughout his capital trial because his attorney was laboring

under a conflict of interest; and (2) whether Dallas received ineffective assistance

of counsel at the penalty phase of his capital trial because his attorneys failed to

adequately investigate and present mitigating evidence. The state court’s

determination that Dallas’s counsel was not encumbered by an actual conflict that

adversely affected his performance was neither contrary to nor an unreasonable

application of clearly established law; nor was it based on an unreasonable

determination of the facts. Nor, finally, has the petitioner established ineffective

assistance of counsel at the penalty phase. We affirm the judgment of the district

court and deny the petitioner habeas relief.

                                           I.

      At trial, overwhelming evidence, including the testimony of Dallas himself,

detailed the abduction and brutal murder of Hazel Liveoak. In addition to the basic


                                           3
              Case: 17-14570     Date Filed: 07/13/2020     Page: 4 of 56



facts we’ve recounted, the jury heard the gruesome details surrounding Liveoak’s

death. The evidence suggested she lived for a number of hours in the trunk of her

car, baking under the hot Alabama sun in July. The state medical examiner

testified that the autopsy he performed revealed that Mrs. Liveoak had bruising on

the right side of her head, the backs of both hands and wrists, and her right bicep,

as well as cuts on her palms, all of which were consistent with her banging on the

trunk lid to get out or call for help. He also observed that Liveoak had urinated

while confined in the trunk of the car. The medical examiner determined her death

by heart attack to be a homicide because, while she was functioning in her daily

life despite her heart disease, “she did not have the cardiac reserve to handle [the]

extremely stressful confines she was in, in a dark, hot, confined trunk of a car and

left there for hours, that her heart could not take that amount of stress.”

      Moreover, according to the testimony of Dennis (“Tony”) Bowen, an

acquaintance of Dallas’s and Yaw’s, the two were bragging at the crack house

about their crime, explaining that they had left an old lady in the trunk of a car.

Bowen added that when he asked Dallas about it, Dallas said that he “hoped the

old lady would die.” The state also presented evidence at trial that Dallas had

abducted and robbed another elderly person, 80-year-old Wesley Portwood, from a

shopping center parking lot in Prattville just three days before kidnapping Mrs.

Liveoak. Portwood testified at trial. He said Dallas abducted him at knifepoint in


                                           4
              Case: 17-14570     Date Filed: 07/13/2020    Page: 5 of 56



his vehicle, drove to a remote area, and ordered Portwood out of the car. Dallas

told Portwood to lie down in the woods, and when Portwood questioned Dallas,

Dallas said he could either lay in the woods or be locked in the trunk of the car

instead. Portwood told Dallas that it was too hot to get into the trunk of the car and

that he would “smother to death in there.” Portwood chose instead to lay in the

woods and survived the robbery and abduction.

                                          II.

      A. Pretrial Appointment of Counsel

      On February 1, 1995, Algert Agricola was appointed by a Montgomery

County Circuit Court judge to represent Dallas. The same day, Agricola was

separately appointed by Alabama’s attorney general for the limited purpose of

representing the Alabama Department of Mental Health and Mental Retardation in

an unrelated civil case in the Middle District of Alabama. The following month,

Agricola moved to withdraw as Dallas’s counsel. Agricola explained that because

of his appointment as a deputy attorney general, he was “subject to the authority of

the Alabama Attorney General who [would] represent the State of Alabama on

appeal from a conviction in [Dallas’s case].” Agricola told the state trial court that

he had conferred with the Disciplinary Commission of the Alabama State Bar

Association and had been advised that the Commission “[did] not believe there

exist[ed] a conflict under the Rules of Professional Conduct in [those]


                                          5
              Case: 17-14570     Date Filed: 07/13/2020    Page: 6 of 56



circumstances,” but Agricola maintained nevertheless that “the question of an

ethical conflict [was] entirely different from the question of whether [Dallas’s]

constitutional rights [were] violated by his being forced to accept the

representation by appointment of an attorney who also serve[d] as a Deputy

Attorney General.” He also submitted an affidavit explaining that it was his

opinion that Dallas’s constitutional rights would be violated by his appointment.

The trial court conducted a hearing and then denied the motion.

      At the hearing, Agricola “represent[ed] to the Court that if [he] were to stay

in this case that [he] would do [his] best for the defendant,” but that he believed

“the test [for conflict] [was] whether there [was] some potential that there might be

duality of representation.” He said he had discussed the matter with Dallas and

Dallas did not want his representation because of his concurrent status as a deputy

attorney general. After Agricola’s brief presentation, the trial judge rejected the

claim of conflict this way: “I know the argument you are making, but I don’t think

there is sufficient conflict of interest to prevent you from representing him. I am

going to follow the response of the Alabama Bar Association until this matter is

concluded.” Agricola, along with co-counsel Jeffery Duffey, who was appointed

in September 1995, represented Dallas throughout the trial. Agricola was

eventually replaced by appellate counsel on direct appeal.

      B. The Guilt Phase Evidence Relevant to Mitigation


                                          6
              Case: 17-14570    Date Filed: 07/13/2020   Page: 7 of 56



      Counsel did not deny Dallas’s involvement in the crime but focused instead

on his intent. In his opening statement, Agricola summed up the defense this way:

      You have to know what happened to [Dallas] on that day and in the days
      before that day, and you need to know about his entire life, because the first
      tragedy in this case is clearly the death of Hazel Liveoak. But the second
      tragedy in this case is the life of Donald Dallas.

      Donald Dallas came from a broken home. You will hear that his parents were
      divorced when he was about six years old. His father beat him. He went to
      bars with his father before he was ten years old. He was ingesting
      intravenously in his veins crystal methamphetamine when he was eleven years
      old. He has constantly used drugs ever since. That is the life Donald Dallas
      has had.

      At the guilt phase, the defense called four witnesses: Rhonda Sue Chavers,

Dr. Guy Renfro, Susan James, and Donald Dallas himself. We detail this evidence

at some length because it formed much (although not all) of the defendant’s

mitigation evidence at the penalty phase.

      Rhonda Sue Chavers, a friend of Dallas’s since he was a teenager, testified

that she was with Dallas on the morning of July 14, 1994. Dallas and Yaw had

slept at Chavers’s house on the night of the 13th, the day after Mrs. Liveoak’s

kidnapping. Chavers said she and Dallas were watching television in the morning

on the 14th and learned that Mrs. Liveoak had died. Chavers testified that Dallas

“cried” and “was worried” when they learned of her death, and told Chavers “that

he tried to get [an acquaintance] to take him back over there, but [he] wouldn’t.”




                                            7
                 Case: 17-14570    Date Filed: 07/13/2020   Page: 8 of 56



          The defense also called Dr. Guy Renfro, a licensed clinical psychologist in

private practice who, after being appointed by the state trial judge, evaluated

Dallas for competency to stand trial. Renfro had a contract with the Taylor Hardin

Secure Medical Facility, which was affiliated with the state’s department of mental

health. Renfro explained that his evaluation of Dallas included Dallas’s mental

state at the time of trial and at the time of the murder. Renfro testified that this

assessment required him to review information about the crime, meet with Dallas,

administer various psychological tests, and obtain information about Dallas’s

background. He told the jury that Dallas had a very long history of substance

abuse, beginning with the consumption of alcohol at age seven or eight, marijuana

at twelve or thirteen, and the intravenous use of crystal methamphetamine and

other addictive drugs at thirteen or fourteen. According to Dr. Renfro, Dallas’s

drug abuse had been continual since he was twelve or thirteen. Renfro testified

that drug abuse at so young an age makes it more difficult for an addict to quit and

impairs the development of social, problem-solving, and coping skills, making it

even more likely that the user will resort to more drugs and alcohol to manage

stress.

          Turning to Dallas’s drug of choice at the time of the murder, Dr. Renfro

explained that crack cocaine, while not physically addictive, produces intense

psychological addiction and cravings; that users develop a tolerance for the drug,


                                            8
               Case: 17-14570      Date Filed: 07/13/2020   Page: 9 of 56



requiring them to obtain more and more to achieve the same effect; and that this

never-ending cycle of needing more of the drug contributes to the addiction.

Renfro opined that the psychological cravings for the drug will make a user very

uncomfortable, and that if a person wants the drug he will do whatever it takes to

get it. At the time of the homicide, Dallas “was binging on the drug, . . . want[ed]

more and more of the drug, and . . . he was in a process of disregarding a lot of

other circumstances to do whatever he could to obtain money for that drug.”

Renfro said that Dallas and Yaw “were binging on [crack] cocaine for about twelve

days, that [Dallas] was oblivious to time, that he was not even aware that he had

had a birth date during that time, that they were just in the process of acquiring

money to purchase cocaine, using the cocaine, and going out and getting whatever

money he needed to do more cocaine.” Dr. Renfro also explained that “there is

some desperation that builds in when people are using large volumes of cocaine,

the way they are more prone[] to use violence and be more confrontational.” The

primary goal of the crack addict is simply to get more.

       Renfro also testified that Dallas was of below-average intelligence, and that

he was not insane at the time of the crime -- that is, Dallas knew right from wrong.

Renfro told the jury that when Dallas described learning that Mrs. Liveoak had

died, Dallas “became tearful,” “seemed to be remorseful that she had died,” and

said that it was not his intent to kill her.


                                               9
              Case: 17-14570     Date Filed: 07/13/2020   Page: 10 of 56



      The defense mitigation consultant, Susan James, testified that the legal

system treats crack cocaine more seriously than powder cocaine, imposing far

greater penalties for crack cocaine offenses, a “recognition that crack cocaine is a

more dangerous substance than powder cocaine” because of its addictiveness,

popularity, and affordability.

      Finally, Dallas testified. He described at length a deeply abusive childhood

and a thoroughly dysfunctional family. He said he dropped out of school in the

sixth grade and began a pattern of drug abuse at a very early age. His family did

not go to church, he received no guidance on telling right from wrong, and he was

left to his own devices as a kid. Dallas explained that his stepfather would sneak

him drinks of beer and whiskey in bars when he was ten or eleven; that he began

smoking marijuana, frequently, at age nine; and that he skipped school and stole

money from his mother to buy drugs. After moving to Prattville, Alabama, Dallas

began injecting cocaine at thirteen or fourteen when a friend, whose father was an

addict, introduced him to the drug. Around that time, Dallas abused other drugs,

including crystal methamphetamine and Quaaludes. Sometime in 1992, Dallas

discovered crack, which he likened to “mind control” and explained that it “just

tells you to get more, do it again.”

      Dallas said that about two weeks before the crime, he and Yaw began using

crack cocaine, although he had previously quit the drug. By the end of the second


                                          10
              Case: 17-14570     Date Filed: 07/13/2020   Page: 11 of 56



day of their crack binge, Dallas had pawned everything in their home to pay for

crack. Then the two began shoplifting to feed their habit. Dallas explained that

“[c]rack is a drug that you want to get off of. You talk about getting off of it, but

you don’t ever really succeed in it. I feel like there is only two ways to get off

crack, and one is death and the other one is damned near to it.”

      Dallas described the days leading up to the murder of Mrs. Liveoak as being

a time of “drifting” and trading stolen goods for crack. He recounted what

happened when he and Yaw saw Mrs. Liveoak in a shopping center in Prattville.

Dallas pushed her into her car while she was loading her groceries, and then drove

off. Liveoak was “scared” at first, but as they were driving, the two began to talk.

He told Liveoak he was a crack addict, and she asked if he was a Christian,

whether he “believe[d] in the Lord,” and if he wanted help. She then prayed with

Dallas in the car. Dallas drove to the end of a road and told her to walk into the

woods. Liveoak expressed fear of the woods, so Dallas suggested that she get into

the trunk. She asked him if she would “be all right in there”; Dallas said she

would, adding that he had ridden in the trunk frequently while attending concerts.

He said he didn’t force her into the trunk because he “didn’t have to. She was a

real nice lady.”

      Dallas and Yaw drove to an ATM machine and used Mrs. Liveoak’s credit

card to withdraw $800. Then they called a cab, leaving Mrs. Liveoak in the trunk


                                          11
             Case: 17-14570      Date Filed: 07/13/2020   Page: 12 of 56



of her car. Liveoak gave Dallas her son’s telephone number, but he didn’t write it

down. He reassured her he would call the police, but he never did. Instead, he

went to purchase more crack and smoke it. Dallas passed many payphones on his

way from the parking lot to a crack house. And when asked why he never called,

Dallas testified that after smoking crack in a motel, he asked his crack dealer,

Chester, if he knew anyone with a car who could take Dallas back to the parking

lot where he had left Liveoak. Dallas said he did not want to call a cab because, in

his words, he “didn’t want to get caught.” Dallas offered a friend of Chester’s $25

to take him back to the parking lot. They started to drive there, but Dallas said that

the car overheated and the two returned to the motel. Dallas told the jury he never

intended to kill Mrs. Liveoak.

      In closing argument at the guilt phase, Agricola told the jury:

      Now, you may say, well, sure, he made that choice, and he did. But, you
      know, you have heard about his life, and you have heard about his stepfather
      giving him drinks when he was six or seven years old, and you heard about
      the fact that he never had anybody telling him what he was doing was wrong
      or showing him what the right thing to do was. He never went to church. He
      was injecting intravenously drugs by the time he was eleven years old.

      It is hard. It is hard for us to think about how a person with that kind of
      upbringing views choices and consequences of those choices . . . .

Agricola also reminded the jury that

      None of you have had an experience of even a few minutes of a life like
      Donald Dallas’, so you don’t know what it is like to live that way. How do
      we know, you know, about sleeping in abandoned houses, about grabbing a
      TV in Wal-Mart and running out and throwing it in the back of a truck so you
                                          12
             Case: 17-14570     Date Filed: 07/13/2020    Page: 13 of 56



      can go get money to go get some crack cocaine. We don’t know anything
      about that.

      On October 19, 1995, the Alabama jury found Donald Dallas guilty of two

capital offenses: murder in the course of a kidnapping in the first degree, in

violation of section 13A-5-40(a)(1) of the Alabama Code; and murder in the course

of a robbery in the first degree, in violation of section 13A-5-40(A)(2) of the

Alabama Code.

      C. The Penalty Phase Evidence

      At the penalty phase, the state offered four aggravating factors: capital

murder in the course of a robbery, Ala. Code § 13A-5-49(4) (1975); capital murder

in the course of a kidnapping, id.; prior crimes of violence (specifically the

kidnapping and robbery of Portwood),

and, finally, the murder

was especially heinous, atrocious, or cruel,

The state relied on

the evidence it had introduced at the guilt phase in support of each of the

aggravators except the conviction for prior crimes of violence. As for prior crimes

of violence, the state introduced certified copies of Dallas’s convictions for

kidnapping in the second degree, robbery in the first degree, robbery in the second

degree, and burglary in the first degree. The state introduced no new evidence

addressing any of the other aggravating factors. It called only one witness, Larry

Liveoak, the victim’s son. Larry testified about the search for his mother, the

profound impact her death had on their family, and his mother’s sterling character.
                                          13
             Case: 17-14570     Date Filed: 07/13/2020   Page: 14 of 56



      The defense claimed three statutory mitigating factors: Dallas was under the

influence of extreme mental or emotional disturbance at the time of the murder,

Ala. Code § 13A-5-51(2); he was under the substantial domination of another

person (Carolyn “Polly” Yaw),

and he lacked the substantial

capacity to conform his conduct to the requirements of the law because of his drug

addiction,

Defense counsel told the jury it would present

testimony about Dallas’s good character along with additional evidence about

Dallas’s tortured life. In his opening statement at the penalty phase, Agricola told

the jury, “[t]o a certain extent, you have heard a lot of this already. But we want

you to understand some further details from members of his family who are here

and who will testify.”

      The defense called four mitigation witnesses: Dallas’s older siblings, Cindy

Knight (now, Cindy Dallas) and Paul Dallas; Dallas’s former partner and mother of

two of his children, Pam Cripple; and Rhonda Lee Chavers, a friend of Dallas and

Yaw who had also testified at the guilt phase. Again, the object of the defense was

to amplify and highlight the shattered life Dallas experienced as a child and as a

young man.

      Cindy Knight, Dallas’s sister, testified she and her brothers were raised in

the same household. She explained that they had an older brother, Jimmy Dallas,

but that he lived with their grandparents. She said their home was filled with


                                         14
             Case: 17-14570    Date Filed: 07/13/2020   Page: 15 of 56



violence, punctuated by the use of knives and guns by their parents. All of the

children, including Donald, saw this pattern of pervasive violence. Knight testified

she had seen her parents chase each other with knives. She was the target of

beatings by her mother, which Donald also witnessed as a young child. She

explained that the violence directed against her and her siblings came at the hands

of their mother. Their father was mostly absent, “either driving a truck or drunk.”

After their parents split up, her mother entered a common-law marriage with

another man, who lived with her and the children in New York, then in Florida,

and finally in Alabama. When asked to describe what it was like growing up in

that household, Knight said simply, “[i]t was hell.” Knight added that her mother

and stepfather “didn’t pay much attention to” Donald and that “[n]obody cared.”

Dallas’s counsel asked Knight whether she could identify any specific instance of

violence that occurred in the home that Donald would have known about; she

responded, “[w]hen I was molested.” She ran away from home at eighteen to

escape.

      Knight further testified that the police and an ambulance had come to their

home and taken their mother “to an insane asylum a couple of times.” She added

that she and Donald were introduced to bars at a very young age: “[w]e have

always been in them. We was raised in them.” The children often went without

food, they were “lucky [they] got food sometimes,” and “[s]ometimes it would be a


                                         15
              Case: 17-14570     Date Filed: 07/13/2020     Page: 16 of 56



week before [they] would get to eat.” Knight said Dallas did not do well in school

-- that is, “whenever [they] got to go to school . . . . if [their mother] got [them] up

or if [they] had clothes clean to wear or whatever.”

      Dallas was never in trouble when he lived with his first common-law wife,

Pam Cripple, but his behavior changed markedly when he became involved with

Yaw. Before Yaw, Dallas “was a kind, considerate person” who “would do

anything for anybody,” but “when he got with [Yaw], he got with the drugs, and it

was like a different person.” “You didn’t know him.” Cindy explained that Yaw

“dominated” Dallas and he would generally take the blame for her. She opined

that it was not in Dallas’s character to intend to kill anyone, but that Yaw had been

involved earlier in an attempted murder.

      Paul Dallas, Donald’s older brother, also testified on his behalf. Counsel

asked him if he “ha[d] any differences” with Cindy Knight’s characterization of

their turbulent home life. “No, sir,” he replied. Paul also discussed their older

brother Jimmy, who was raised by their grandparents and went on to lead a

successful life, having gone to college and worked as a counselor. Paul too

described Donald’s good character before he met Yaw. Paul said Donald “would

do anything for people,” “work on their houses when they couldn’t get nobody to

do it and didn’t have the money,” “fixed people’s cars,” and “went to trailer parks

and cut grass for the old people.” But after Donald met Yaw, “he just quit and


                                           16
               Case: 17-14570   Date Filed: 07/13/2020    Page: 17 of 56



started running around with her and her buddies.” In the weeks leading up to the

murder, Dallas did not work; instead, he was “[o]ff running around doing crack

with Polly.”

      Pam Cripple, Dallas’s first common-law wife and mother of his two eldest

children, testified that when she lived with Dallas, he was “kind” and tended to

help others. Dallas and Cripple split up when Dallas began seeing Yaw. Cripple

said that Dallas changed when he met Yaw, and then “had very little to do with his

kids.” She said her children had been in contact with Dallas since the murder and

they wanted to continue to see him. She also said it was out of character for Dallas

to kill anyone; in fact, he was never known to be violent.

      Finally, Rhonda Chavers testified that she knew both Dallas and Yaw well,

that Dallas was not a violent person, but that Yaw was “mean” and controlled him.

Chavers reiterated her earlier testimony that Dallas cried when he learned that Mrs.

Liveoak had died and that he was greatly affected by her death.

      In closing argument, the prosecutor recounted the trial evidence and the

most salient aggravating factor -- that the murder was especially heinous,

atrocious, or cruel. The jury was reminded: “Hazel Liveoak died as gruesome a

death as you can imagine,” “entombed in her car,” given false hope that the

defendant would call for help, but instead she was left alone in a “stuffy, hot trunk

fighting for her life, fighting for a chance, a chance that never came to her.” The


                                          17
             Case: 17-14570     Date Filed: 07/13/2020   Page: 18 of 56



prosecutor argued that it would have been less cruel if Dallas had shot or stabbed

his victim; instead “he chose the two cruelest weapons that he could choose, time

and terror.” Ultimately, he argued that “no matter how many mitigating

circumstances [the defense] could put forth of how bad his life has been, they

simply do not, cannot outweigh what he did to Hazel Liveoak.”

      In Dallas’s defense, Agricola again asked the jury to reflect on Dallas’s life:

      What kind of people would any of us be if we had the same kind of start in
      life that Donald Dallas had? What if we had been shooting drugs in our veins
      at age twelve? What if we didn’t have anybody that cared about us as a child?
      Children aren’t born bad. Something happens to them that makes them bad. .
      . . Donald Dallas’ world is not our world, never has been.

      The same day, the jury recommended by a vote of 11 to 1 that Donald Dallas

be sentenced to death. The trial court followed the jury’s recommendation and

sentenced Dallas to death on November 16, 1995.

      D. Direct Appeal

      On direct appeal, the Alabama Court of Criminal Appeals allowed trial

counsel Agricola to withdraw. New counsel was appointed but Jeffery Duffey

remained on the case. Dallas’s convictions and sentence were affirmed on direct

appeal. Dallas v. State, 

(Ala. Crim. App. 1997). Dallas argued,

among other things, that his lead trial counsel, Agricola, rendered ineffective

assistance because he was encumbered by a conflict of interest. Applying Cuyler

v. Sullivan, 

(1980), the Court of Criminal Appeals rejected the claim,


                                         18
                 Case: 17-14570        Date Filed: 07/13/2020        Page: 19 of 56



reasoning that Dallas had failed to establish an “actual conflict of interest” because

Dallas’s “interests were not adverse to that of the Alabama Department of Mental

Health and Mental Retardation.” He failed to “establish[] by any evidence that his

attorney’s performance was adversely affected” by the alleged conflict, and he

failed to show prejudice. 

. The state appellate court

squarely rejected Dallas’s argument that prejudice should be presumed in this case

because Dallas failed to establish an actual conflict. 1

summary order, the


1
    The Alabama Court of Criminal Appeals explained its ruling this way:

         The appellant argues that he received ineffective assistance of counsel, in violation of the
         Sixth and Fourteenth Amendments to the United States Constitution, because of, he says,
         an actual conflict of interest on the part of his trial counsel.

         A review of the record indicates that the appellant’s argument is without merit. Before
         trial, one of the appellant’s attorneys filed a motion to withdraw from representation of the
         appellant. The motion was based upon that attorney’s having been previously appointed
         by the attorney general to represent the Alabama Department of Mental Health and Mental
         Retardation in an unrelated civil case. The attorney stated in his motion that he had
         conferred with the Disciplinary Commission of the Alabama State Bar Association, which
         issued an advisory opinion stating that a conflict did not exist. The trial court denied the
         motion without a hearing.

         “In order to demonstrate a violation of his Sixth Amendment rights, a defendant must show
         that an actual conflict of interest adversely affected his lawyer’s performance.” Cuyler v.
         Sullivan, 

, 350 (1980). “An actual conflict of interest exists when an attorney
         owes loyalty to a client whose interests are adverse to another client.” Self v. State, 

, 1033 (Ala. Crim. App. 1989), cert. quashed, 

(Ala. 1990).

         While it is true that the attorney general had appointed the appellant’s trial counsel as a
         deputy attorney general in an unrelated civil matter, an actual conflict of interest did not
         exist in this case because the appellant’s interests were not adverse to that of the Alabama
         Department of Mental Health and Mental Retardation. The appellant argues that we must
         presume prejudice is in this case because the attorney had to struggle to serve two different
         “masters.” On the one hand he was commissioned as a deputy attorney general and in that
         capacity his superior was the attorney general. On the other hand, he was representing a


                                                  19
               Case: 17-14570        Date Filed: 07/13/2020       Page: 20 of 56



Alabama Supreme Court affirmed, Ex parte Dallas, 

(Ala. 1998),

and the Supreme Court denied certiorari, Dallas v. Alabama, 

(1998).

       E. State Postconviction Proceeding

       Dallas sought collateral review in a Rule 32 petition filed in the

Montgomery County Circuit Court. He raised a variety of claims, including the

second claim now before us -- that his counsel was ineffective at the penalty phase

because they did not locate or call several witnesses who could have offered

additional mitigating evidence. Trial counsel Jeffery Duffey and

investigator/mitigation consultant Susan James, testified, and the court accepted a

deposition from trial counsel Algert Agricola. On September 25, 2001, after the

hearing, the court denied the ineffective assistance claim.




       defendant charged with a crime that the attorney general is charged with enforcing.
       However, prejudice is presumed only when an actual conflict is shown. 

–50. See also Browning v. State, 

, 342 (Ala. Crim. App. 1992).
       Because no actual conflict of interest existed in this case, prejudice is not presumed.

       Additionally, the appellant has not demonstrated that he was prejudiced as a result of the
       fact that his attorney represented the State in the civil case. Nor has the appellant
       established by any evidence that his attorney’s performance was adversely affected by his
       appointment in that case. The appellant must make a factual showing that his attorney
       “made a choice between possible alternative courses of action, such as eliciting (or failing
       to elicit) evidence helpful to one client but harmful to the other.” Self v. 

.
       (citations omitted.)

       Therefore, the appellant was not denied effective assistance of counsel or a fair trial
       because of an alleged conflict of interest on the part of his trial counsel.

.
                                               20
             Case: 17-14570     Date Filed: 07/13/2020    Page: 21 of 56



      First, the circuit court summarily dismissed the ineffectiveness claim based

on conflict of interest because the claim had been raised and rejected on direct

appeal by the Alabama Court of Criminal Appeals. As for the penalty-phase

ineffectiveness claim, the trial court concluded that it had been abandoned because

Dallas failed to call any of the relevant witnesses at his Rule 32 hearing.

Alternatively, it determined that Dallas’s penalty-phase ineffectiveness claim was

without merit. “Though the alleged testimony of the additional witnesses may

have provided specific instances of abuse, neglect, or drug abuse in Dallas’

childhood and adult life, the sum total of the testimony would have been, at best,

cumulative and would not have changed the outcome of sentencing.” In short,

“Dallas failed to show how trial counsel’s performance was deficient, or that he

suffered any prejudice.”

      A month later, Dallas filed a motion to alter, vacate, or amend that order.

Four days later, the circuit court denied the application. But Dallas did not file his

notice of appeal until November 28, 2001, weeks after Alabama’s 42-day

jurisdictional window to file an appeal had closed. The Alabama Court of

Criminal Appeals granted the state’s motion to dismiss the appeal as untimely. In

February 2001, Dallas moved the circuit court for an order holding that his filing of

a motion to vacate, alter, or amend the court’s denial of his Rule 32 petition tolled

the statutory deadline for filing an appeal. The circuit court granted the motion.


                                          21
              Case: 17-14570     Date Filed: 07/13/2020    Page: 22 of 56



When Dallas filed a second notice of appeal with the Alabama Court of Criminal

Appeals two weeks later, the state moved to strike the notice. The court agreed,

granted the motion and struck the second notice of appeal, observing that it had

already dismissed his earlier appeal for untimeliness. The Alabama Supreme

Court denied certiorari.

      F. Federal Habeas Proceedings

      Dallas then turned his sights to the federal district court, filing this § 2254

petition in the United States District Court for the Middle District of Alabama.

The court rejected both claims now before this Court: whether defense counsel

labored under a conflict of interest in violation of the Sixth Amendment; and

whether defense counsel was ineffective at the penalty phase by failing to conduct

a reasonable mitigation investigation and by failing to present additional mitigating

evidence. As for the alleged conflict, the district court found that the state court’s

denial of the claim was neither contrary to nor an unreasonable application of

clearly established federal law; nor was it based on an unreasonable determination

of the facts in light of the evidence presented.

      When it considered the ineffective-assistance-of-counsel claim, the district

court avoided the procedural default issues and reviewed the claim de novo rather

than affording deference to the state court’s determination under the Antiterrorism

and Effective Death Penalty Act (“AEDPA”). It was free to do so under


                                          22
              Case: 17-14570     Date Filed: 07/13/2020    Page: 23 of 56



controlling case law. Loggins v. Thomas, 

, 1215 (11th Cir. 2011)

(“When relief is due to be denied even if claims are not procedurally barred, we

can skip over the procedural bar issues, and we have done so in the past.”);

see Bell v. Cone, 
543 U.S. 447
, 451 n.3 (2005) (per curiam)

(citing § 2254(b)(2) for the proposition that “an application for habeas corpus may

be denied on the merits, notwithstanding a petitioner’s failure to exhaust in state

court”). The district court also considered only the prejudice prong of Strickland v.

Washington, 
466 U.S. 668 
(1984), which it was likewise free to do. Waters v.

Thomas, 
46 F.3d 1506
, 1510 (11th Cir. 1995) (explaining that a “court may decline

to reach the performance prong of the ineffective assistance test if convinced that

the prejudice prong cannot be satisfied” (citing 
Strickland, 466 U.S. at 697
)). The

court concluded that Dallas had failed to show prejudice because there was “no

reasonable probability that, but for any of the acts or omissions of Petitioner’s trial

counsel identified by Petitioner in his rambling federal habeas corpus pleadings

and briefs, the outcome of either phase of Petitioner’s 1995 capital murder trial

would have been any different.”

      G. Additional Mitigating Evidence Presented to the Federal Habeas Court

      The new mitigating evidence considered by the district court was presented

in affidavits submitted in 2007. Most critical among the new evidence were

additional averments from Cindy and Paul Dallas, as well as from Dallas’s mother,


                                          23
             Case: 17-14570      Date Filed: 07/13/2020   Page: 24 of 56



Elaine Dallas, his other brother, Jimmy Dallas, and an expert opinion offered by a

new psychologist, Dr. Ken Benedict. We recite this evidence too in some detail in

order to properly evaluate the petitioner’s claim that his counsel were ineffective

for failing to present additional mitigation at the penalty phase.

      James (“Jimmy”) Dallas, Donald’s older brother who had not testified at the

trial, submitted an affidavit attesting to the following. The children’s father, James

Dallas, Sr., was an alcoholic who was largely absent, and much of his time at home

was spent fighting with their mother, Elaine. Because of his parents’ infidelity,

there were many discussions at home about whether Donald was in fact James,

Sr.’s son. Jimmy also testified that the children were often left home alone while

their parents were out drinking. He confirmed that their mother was unstable,

abused drugs, and was violent. She once threatened to kill all of the children with

an eight-to-ten-inch butcher knife while their father was out. Because of her

mental illness, Jimmy explained, their mother did not take good care of the

children. There were stray cats and dogs living in their home -- up to twenty at one

point -- and the home smelled terribly. Jimmy described living with his mother as

“torture.” He said his father moved in and out of their home, and when their

parents eventually divorced, his mother started up with a man named Chick, “who

was also a drunk.”




                                          24
              Case: 17-14570      Date Filed: 07/13/2020    Page: 25 of 56



      Because he was an athlete in high school, Jimmy had coaches and other

positive role models in his life, whereas Donald lacked any role models. Even so,

Jimmy had a drug and alcohol problem until finally entering Alcoholics

Anonymous at the age of 29. As a result of counseling, he confronted many of the

negative aspects of his upbringing, including that he and his siblings were never

taught right from wrong, they weren’t taught how to care for themselves, their

parents lacked money to pay for medical or dental care for them, they frequently

lacked food and clothing, and he and his siblings all experienced significant

problems in their lives as a result of their troubled childhood. He offered the view

that his siblings “really never had a chance in life”; while Jimmy had other positive

influences, school, sports, and the military, his siblings “had no alternative at all.”

      Donald’s sister Cindy, who testified at trial, also submitted an affidavit in

2007. She averred that their father James “was a violent, mean man who was a

drunk and messed around with other women.” The children were often left alone

while their parents were out drinking. Cindy elaborated on the fighting between

their parents, testifying that their father beat their mother in front of the children,

chased her around the house with knives and guns, threatening to kill her, and that

at times the children believed he would. She said their father took the children

with him to bars, where sometimes “bartenders fed [the kids] so [they] would not

go hungry.” Their father also took them to the homes of women with whom he


                                           25
              Case: 17-14570     Date Filed: 07/13/2020    Page: 26 of 56



was cheating. Cindy believed her father’s behavior “actually drove [their] mother

crazy” and explained that she was twice “[taken] . . . away in a straight jacket.”

Cindy testified that her mother was “sick,” that she “would see things that [others]

could not see,” and, at times, made the children hide as a result. Until their parents

broke up, her mother would frequently beat Cindy. Because they were so poor,

they often lacked food and clothing, and they moved around a lot between “houses

that were in bad shape,” including one that was “full of rats.” Donald was once

bitten by a rat and taken to the hospital. Cindy said that after their parents

divorced, their mother began a relationship with Chesley (“Chick”) Collier, who

played in a band, drank a lot, and took Donald to play the drums in bars when he

was eight or nine, but Chick “was good to [Paul and Donald].”

      Cindy testified that Donald began a relationship with Pam Cripple when the

pair were teenagers, and that when Donald was with Pam, he worked full-time,

supported his family, and took care of their mother, Elaine. But Donald changed

dramatically after he left Pam and began a relationship with Polly Yaw. He started

using drugs. While Dallas attempted to maintain a relationship with his daughters,

Pam did not want Yaw around the kids. Cindy testified that Yaw was quick-

tempered and physically abusive and controlling of Donald. Yaw and Donald had

four children together, plus Yaw’s oldest daughter who Donald helped raise. She




                                          26
               Case: 17-14570     Date Filed: 07/13/2020    Page: 27 of 56



added that Yaw would manipulate Donald by threatening to take the children

away.

        Paul Dallas, who had also testified at trial, likewise submitted an affidavit in

2007. Like Cindy and Jimmy, Paul testified that their parents were heavy drinkers,

fought, and were abusive of each other and the children. His father spent all of his

money on alcohol and the children often went hungry. Paul said his mother did her

best to support them, sometimes working two jobs, but the family remained so

poor that when the children’s shoes wore down, their mother put cardboard in them

because they couldn’t afford new ones. Paul testified just like Cindy had regarding

their rat-infested homes. He explained that when Donald was between twelve and

fourteen, he began cutting school, drinking alcohol, and smoking cigarettes and

marijuana. Paul also testified about his mother’s relationship with Chick and his

influence on the children. He took them to bars, and he was “good to [them] but he

drank a lot.” Paul said they spent a lot of time with Chick’s sister, Patricia

Mefford, and one of Ms. Mefford’s male friends molested both Paul and Donald.

Paul explained that he “witnessed Donald being anally raped as well as being

forced to perform oral sex on this man” and that “Donald also witnessed the same

thing happen to” Paul. This happened on at least four occasions. However, he did

not identify the name of the abuser, the time, or the location.




                                           27
             Case: 17-14570     Date Filed: 07/13/2020   Page: 28 of 56



      Donald’s mother, Elaine, also submitted an affidavit. She said that her

relationship with Donald’s father was abusive, and that he beat her severely, even

when she was pregnant. James, Sr. was abusive to the children as well, once

throwing Paul against a wall. She smoked and drank throughout her pregnancies,

consuming four or five vodka cocktails a day. She testified about the family’s

poverty and her inability to pay for certain medical care. She also testified about

her severe chronic depression and hallucinations. She was admitted to a

psychiatric hospital at least once, where she remained for about a month until her

husband forced her to sign herself out against the recommendation of her doctors.

      Elaine admitted in her affidavit that as her drinking increased, she became

more abusive to her children, spanking them with a paddle. She confirmed the

details of her relationship with Chick and his influence on her children. She

became aware that Paul and Donald skipped school and used alcohol and drugs,

and although she spoke to the school’s truancy official about the problem, the

school did nothing. Ultimately, Elaine Dallas offered the following:

      It is my opinion that I put my children through hell when they were growing
      up. I exposed them to alcohol abuse by their father and by me. They were
      exposed to physical abuse by their father and by me. They witnessed their
      father beating me on many, many occasions. They were around me when I
      was out of my head and I am certain they must have been afraid of my
      behavior. My children were also forced to live in poverty and as a result they
      had very little security when they were growing up.




                                         28
              Case: 17-14570     Date Filed: 07/13/2020    Page: 29 of 56



      Dr. Ken Benedict, a psychologist who evaluated Donald, also submitted an

affidavit. Benedict spent eight-and-a-half hours with Dallas and two-and-a-half

hours reviewing available records, including affidavits from his family and others

familiar with the Dallas family. Benedict described Dallas as having “average

intellectual ability,” though his scores on subsections of the intelligence test varied

widely, “making it difficult to summarize his general intelligence with one score.”

Although Dallas had improved in reading and spelling while incarcerated, Dallas

functioned below a fifth-grade reading level, which would be categorized as

“impaired.” Dr. Benedict diagnosed three learning disorders (in reading, written

language, and mathematics), as well as Attention Deficit Hyperactivity Disorder

(“ADHD”). The learning disorders were akin to severe dyslexia, and the ADHD

was characterized as having difficulty with sustained attention, impulse control,

and planning, among other impairments. He explained that these conditions could

be remediated or exacerbated by environmental factors, that Dallas’s conditions

were not identified or consistently treated, and that there were highly negative

psychosocial influences in his family life, including poverty, neglect, physical and

sexual abuse, traumatic exposure and substance abuse.

      Benedict opined that people with Dallas’s learning and attention disorders,

particularly when untreated, are more likely to develop substance abuse disorders

during early adolescence. Because of the hardship and shame of having learning


                                          29
             Case: 17-14570     Date Filed: 07/13/2020    Page: 30 of 56



disorders, negative feedback from teachers and peers, and the lack of parental

oversight, such individuals frequently drop out of school. Benedict added that

Dallas suffers from depression, anxiety, low self-esteem, shame, and cognitive

difficulties in the form of inattention, disorganization, and confusion under stress.

People with these characteristics are often described as unassertive or passive-

dependent. Moreover, people who share Dallas’s psychological profile “are

typically deficient in executive brain functions” and get into trouble before having

time to plan or reflect on their behavior. Benedict said that Dallas “does not fit the

personality profile of someone who would act solely or take the lead in a crime

such as the one in question”; he would be more likely to assume the role of a

“follower.” Finally, Benedict offered his opinion that “it is not likely Mr. Dallas

desired or planned the death of the victim.”

      In conducting de novo review of the ineffectiveness claim, the district court

described the defendant’s mitigation case at trial as being “substantial.” Moreover,

he observed that in some ways the new evidence would have weakened the trial

presentation by providing conflicting accounts of Dallas’s mother’s mental health

for example, and in other ways merely duplicated what the jury had heard.

Weighing the totality of the old and new mitigating evidence against the powerful

aggravators presented by the state, the district court concluded that there was no




                                          30
                 Case: 17-14570        Date Filed: 07/13/2020       Page: 31 of 56



reasonable probability that the outcome would have been any different if the jury

had heard it all, and it denied the claim.

                                                 III.

         “We review de novo a district court’s grant or denial of a habeas corpus

petition.” McNair v. Campbell, 
416 F.3d 1291
, 1297 (11th Cir. 2005). Because

Dallas filed his federal habeas petition after April 24, 1996, this case is governed

by AEDPA. “Under AEDPA, if a state court has adjudicated the merits of a claim

-- as the state court did here -- we cannot grant habeas relief unless the state court’s

decision ‘was contrary to, or involved an unreasonable application of, clearly

established Federal law, as determined by the Supreme Court of the United States,’

or ‘was based on an unreasonable determination of the facts in light of the

evidence presented in the State court proceeding.’” Kilgore v. Sec’y, Fla. Dep’t of

Corr., 
805 F.3d 1301
, 1309 (11th Cir. 2015) (quoting 28 U.S.C. § 2254(d)).2



2
    Pursuant to § 2254(d):

         An application for a writ of habeas corpus on behalf of a person in custody pursuant to
         the judgment of a State court shall not be granted with respect to any claim that was
         adjudicated on the merits in State court proceedings unless the adjudication of the claim
         --

             (1) resulted in a decision that was contrary to, or involved an unreasonable
                 application of, clearly established Federal law, as determined by the Supreme
                 Court of the United States; or

             (2) resulted in a decision that was based on an unreasonable determination of the
                 facts in light of the evidence presented in the State court proceeding.


                                                  31
              Case: 17-14570     Date Filed: 07/13/2020    Page: 32 of 56



      “Under § 2254(d)(1)’s ‘contrary to’ clause, we grant relief only ‘if the state

court arrives at a conclusion opposite to that reached by [the Supreme] Court on a

question of law or if the state court decides a case differently than [the Supreme

Court] has on a set of materially indistinguishable facts.’” Jones v. GDCP

Warden, 
753 F.3d 1171
, 1182 (11th Cir. 2014) (alterations in original) (quoting

Williams v. Taylor, 
529 U.S. 362
, 413 (2000)). “Under § 2254(d)(1)’s

‘unreasonable application’ clause, we grant relief only ‘if the state court identifies

the correct governing legal principle from [the Supreme] Court’s decisions but

unreasonably applies that principle to the facts of the prisoner’s case.’”
 Id.

(alteration in 
original) (quoting 
Williams, 529 U.S. at 413
).

      The second prong of § 2254(d) -- that an adjudication resulted in a decision

that was based on an unreasonable determination of the facts in light of the

evidence presented in the state court proceeding -- “requires that we accord the

state trial court substantial deference.” Brumfield v. Cain, 
576 U.S. 305
, 314

(2015). “If ‘[r]easonable minds reviewing the record might disagree about the

finding in question, on habeas review that does not suffice to supersede the trial

court’s . . . determination.’”
 Id. (alteration and 
ellipsis in original) (quoting Wood

v. Allen, 
558 U.S. 290
, 301 (2010)).

                                          IV.




                                          32
             Case: 17-14570     Date Filed: 07/13/2020    Page: 33 of 56



      Dallas first claims that his lead trial counsel was ineffective because he was

laboring under a prohibitive conflict of interest. The Sixth Amendment to the

United States Constitution guarantees that “[i]n all criminal prosecutions, the

accused shall enjoy the right . . . to have the Assistance of Counsel for his

defence.” U.S. Const. amend. VI. Assistance of counsel is the bedrock of a fair

criminal justice system. “Thus, the Sixth Amendment does more than require the

States to appoint counsel for indigent defendants. The right to counsel prevents the

States from conducting trials at which persons who face incarceration [or death]

must defend themselves without adequate legal assistance.” Cuyler v. Sullivan,

446 U.S. 335
, 344 (1980). The effective assistance of counsel demands not only a

minimally competent lawyer, but also counsel unburdened by a conflict of interest

that impedes zealous representation.

      Dallas says that because Agricola represented Alabama’s Department of

Mental Health and Mental Retardation (the “DMHMR”) in an unrelated civil

matter, while at the same time defending him against Alabama’s prosecution, he

was deprived of his Sixth Amendment right to the effective assistance of counsel.

He argues that the Alabama Court of Criminal Appeals erroneously applied

Sullivan instead of Holloway v. Arkansas, 
435 U.S. 475 
(1978), in adjudicating the

alleged conflict. In Holloway, the Supreme Court held that “whenever a trial court

improperly requires joint representation over timely objection reversal is


                                          33
             Case: 17-14570     Date Filed: 07/13/2020    Page: 34 of 56



automatic.” 435 U.S. at 488
. The problem with the argument, however, is that

Holloway’s automatic reversal rule is limited only to those circumstances where a

trial court improperly requires the joint representation of codefendants overly

timely objection. In the absence of the joint representation of codefendants, the

appropriate legal standard is found in Sullivan, which requires a showing that an

actual conflict of interest adversely affected defense counsel’s 
performance. 446

U.S. at 349
–50. The Court of Criminal Appeals’ reliance on Sullivan was neither

contrary to nor an unreasonable application of clearly established Supreme Court

law. Moreover, nothing in the state court’s findings amounted to an unreasonable

determination of the facts.

      The typical showing a petitioner must make to succeed on an ineffective

assistance of counsel claim is rigorous. Under Strickland, a petitioner must show

that his counsel’s performance was constitutionally deficient -- that his lawyer

“made errors so serious that counsel was not functioning as the ‘counsel’

guaranteed the defendant by the Sixth Amendment” -- and that the deficient

performance prejudiced the defendant, depriving him “of a fair trial, a trial whose

result is 
reliable.” 466 U.S. at 687
. Simple mistakes or strategic errors are not

enough, nor are serious errors if, absent those errors, there is no “reasonable

probability” that the outcome would have been different.
 Id. at 694. 
But the

Supreme Court has carved out a few “exceptions” to Strickland’s high bar where


                                          34
              Case: 17-14570     Date Filed: 07/13/2020     Page: 35 of 56



counsel’s ineffectiveness stems from a conflict of interest. See Mickens v. Taylor,

535 U.S. 162
, 176 (2002) (describing the precedential line of “exceptions from the

ordinary requirements of Strickland . . . where Strickland itself is evidently

inadequate to assure vindication of the defendant’s Sixth Amendment right to

counsel”).

      In Mickens, the Supreme Court explained the key differences between

Holloway and Sullivan. In analyzing a Sixth Amendment conflict-of-interest

claim, we ask whether a trial court’s duty to investigate an attorney’s conflict of

interest was triggered; and, where a trial court fails to discharge that duty, what is

the remedy. 
Mickens, 535 U.S. at 168
–74. Under Holloway, the duty to inquire is

triggered by a pre-trial objection to the joint representation of codefendants. Under

Sullivan, the duty is triggered, even absent objection, where a trial court knows or

reasonably should know that a particular conflict exists -- whether that conflict

relates to joint representation of codefendants or to other conflicts, like the fee

arrangement in Wood v. Georgia. See 
450 U.S. 261
, 271–73 (1981) (finding that

the trial court had a duty to investigate a conflict resulting from a fee arrangement

between counsel and the defendant’s employer, and remanding for consideration of

whether an actual conflict negatively impacted counsel’s performance).

      If a trial court fails to investigate a conflict that it is obliged to examine, we

address the appropriate remedy. In Mickens, the Supreme Court explained that


                                           35
             Case: 17-14570      Date Filed: 07/13/2020   Page: 36 of 56



Holloway “creates an automatic reversal rule only where defense counsel is forced

to represent codefendants over his timely objection, unless the trial court has

determined that there is no 
conflict.” 535 U.S. at 168 
(emphases added). But

under Sullivan, where a trial court fails to adequately investigate any other type of

conflict that it knows or reasonably should know about, reversal is only warranted

if the petitioner shows an actual conflict that negatively affected his attorney’s

performance. Under Sullivan, then, a defendant is obligated to establish that the

alleged conflict adversely affected his attorney’s performance.

      While Dallas urges us to apply Holloway’s automatic reversal rule to any

alleged conflict just because an objection was made before trial, doing so would

extend Holloway beyond its limited application. What’s more, Dallas asks us to

find, pursuant to our deferential review under AEDPA, that the Alabama Court of

Criminal Appeals’ failure to apply Holloway was contrary to clearly established

Supreme Court law. But the clarity we find in Mickens is just the opposite of what

Dallas has suggested: Holloway “creates an automatic reversal rule only where

defense counsel is forced to represent codefendants over his timely objection.”
 Id.

(emphases added).
      As a preliminary matter, we think it clear that some inquiry was necessary

because the potential conflict was brought to the trial court’s attention by Dallas’s

counsel, who filed a motion to withdraw based on the conflict. We also think it


                                          36
              Case: 17-14570     Date Filed: 07/13/2020    Page: 37 of 56



likely that the trial court adequately discharged its duty. Indeed, the trial court

took argument on Agricola’s motion, listened to everything counsel presented,

considered the ethical opinion of the Disciplinary Commission of the Alabama

State Bar Association and Agricola’s statement that he would “do [his] best” for

his client, and concluded that there was no actual conflict of interest that would

prevent Agricola from vigorously representing Dallas.

      But even if the trial court’s inquiry were somehow deficient, it still was not

an unreasonable application of Sullivan for the Alabama Court of Criminal

Appeals to find that there was no actual conflict of interest that warranted the

reversal of Dallas’s conviction. Dallas’s counsel, Agricola, was appointed as a

Special Deputy Attorney General to represent the DMHMR in a civil class action

relating to conditions in Alabama’s mental health institutions. See Wyatt ex rel.

Rawlins v. Hanan, 
170 F.R.D. 189
, 191 (M.D. Ala. 1995) (describing the multiple

iterations of that litigation, which began in 1970 and resulted in a 1986 consent

decree, compliance with which was litigated again in the 1990s, the point at which

Agricola became involved). That case was completely unrelated to Dallas’s

homicide case.

      But beyond the fact that Agricola represented the DMHMR only in a limited

capacity in a wholly unrelated civil case, the DMHMR’s interests were not adverse

to Dallas’s. First, Dallas says that the DMHMR is directly involved in almost


                                           37
               Case: 17-14570        Date Filed: 07/13/2020       Page: 38 of 56



every capital case through court-ordered psychological evaluations for competency

to stand trial and mental state at the time of the offense. Dallas notes that the trial

court in fact ordered a DMHMR-appointed examination, which was conducted by

Dr. Guy Renfro, who was contracted by the DMHMR. But the DMHMR’s

involvement in contracting for a psychological evaluation of Dallas by a competent

expert does not make its interests adverse to Dallas’s. Dallas has identified no

interest that the DMHMR had in Dallas’s trial beyond the general observation that

it was involved in contracting for the evaluation.3

       Most critically, Dallas has failed to show that this supposed conflict had any

bearing on his counsel’s performance, much less that it amounted to an

unreasonable application of Sullivan or an unreasonable determination of the facts

for the Court of Criminal Appeals to conclude as much. Dallas says that the

alleged conflict caused the defense to rely only on Dr. Renfro as a defense expert,


3
  Dallas relies heavily on Zuck v. Alabama, 
588 F.2d 436 
(5th Cir. 1979), a former Fifth Circuit
case that is binding on this Court. Bonner v. City of Prichard, 
661 F.2d 1206
, 1209 (11th Cir.
1981) (en banc). Zuck is a pre-AEDPA case that neither cited nor interpreted any Supreme
Court law, much less did it tell us what has been clearly established by the Supreme Court in
conflict-of-interest ineffectiveness cases. Indeed, Zuck explicitly did not apply the adverse-
impact requirement of Sullivan, saying instead that the Court’s “analysis in conflict of interest
cases does not focus on the actual effect of the conflict on a particular defendant’s case.” 
Zuck,
588 F.2d at 439
–40. Moreover, Zuck is factually distinguishable from Dallas’s case. In Zuck,
the law firm representing the criminal defendant represented, in an unrelated matter, the state
prosecutor who actually tried Zuck. The Zuck Court noted that “the defense attorneys were
subject to the encumbrance that the prosecutor might take umbrage at a vigorous defense of Zuck
and dispense with the services of their firm.”
 Id. at 439. 
The DMHMR was not a participant in
Dallas’s trial, beyond having contracted for Dallas’s pre-trial evaluation. The claimed conflicts
are not comparable. Whatever we might say about Zuck, the interests of the DMHMR were not
pitted against Dallas’s interests.
                                               38
             Case: 17-14570     Date Filed: 07/13/2020    Page: 39 of 56



rather than hiring a different expert. But there is no reason to think the defense

relied only on and called Dr. Renfro because of Agricola’s alleged conflict. In

fact, in the Rule 32 proceedings, Agricola’s co-counsel, Jeffery Duffey, testified

that they chose to use Dr. Renfro and they weren’t concerned about a conflict

because Renfro was simply contracted by the state in much the same way an

appointed defense counsel is.

      As the state points out, Renfro’s testimony was altogether beneficial to

Dallas; he fully supported the defendant’s theory that he did not intend to kill

Liveoak. As we’ve noted already, Renfro testified that Dallas had a “very long

history of substance abuse” and that “research has shown when people begin drug

abuse at an early age, they have more difficulty later on discontinuing the drug

use,” which “tends to retard their development as far as social development,

problem solving, [and] coping skills.” He also said that when individuals are

“acutely intoxicated,” they have trouble distinguishing right from wrong; that crack

cocaine produces “a very intense psychological addiction with very intense

cravings” and “a lot of discomfort”; and that an addict’s primary goal is to obtain

more and more of the drug. Renfro opined that Dallas was a cocaine addict and

that he was binging on the drug at the time of the homicide. He explained that the

defendant would do whatever he had to do to feed his habit. Dr. Renfro also




                                          39
              Case: 17-14570     Date Filed: 07/13/2020    Page: 40 of 56



testified that Dallas evinced real remorse, becoming “tearful” when he discussed

Liveoak’s death. Dallas told him that it “had not been his intent for her to die.”

      Quite simply, there is no basis in this record to suggest, let alone find, that

the defense team chose to use Dr. Renfro and not another expert because of any

claimed conflict of interest. And this is the only specific, substantive reason

offered in support of the petitioner’s theory that Agricola’s representation was

adversely impacted by his alleged conflict. The state court’s application of

Sullivan to Dallas’s conflict-of-interest ineffectiveness claim was not contrary to

clearly established federal law. And its conclusion that Dallas had failed to show

that his attorney labored under an actual conflict of interest that adversely affected

his attorney’s performance was neither an unreasonable application of Sullivan,

nor was it based on an unreasonable determination of the facts in light of the

evidence presented.

                                          V.

      Dallas also says that he received ineffective assistance of counsel during the

penalty phase of his capital trial because his lawyers failed to adequately

investigate and present mitigating evidence. As we have explained, to succeed on

an ineffective-assistance-of-counsel claim, Dallas must show that: (1) “counsel’s

representation fell below an objective standard of reasonableness,” and (2) “there

is a reasonable probability that, but for counsel’s unprofessional errors, the result


                                          40
              Case: 17-14570     Date Filed: 07/13/2020   Page: 41 of 56



of the proceeding would have been different.” 
Strickland, 466 U.S. at 688
, 694;

accord Knowles v. Mirzayance, 
556 U.S. 111
, 126–27 (2009); Wiggins v. Smith,

539 U.S. 510
, 521–22 (2003); Williams v. Taylor, 
529 U.S. 362
, 390–91 (2000);

Darden v. Wainwright, 
477 U.S. 168
, 184 (1986). “A court may decline to reach

the performance prong of the ineffective assistance test if convinced that the

prejudice prong cannot be satisfied.” 
Waters, 46 F.3d at 1510
. The failure to meet

either Strickland prong is fatal to the claim.

      To show prejudice,

      it must be established that, but for counsel’s unprofessional performance,
      there is a reasonable probability the result of the proceeding would have been
      different. See 
Strickland, 466 U.S. at 694
. “It is not enough for the
      [petitioner] to show the errors had some conceivable effect on the outcome of
      the proceeding . . . ,” because “[v]irtually every act or omission of counsel
      would meet that test.”
 

Nevertheless, a petitioner “need not show
      that counsel’s deficient conduct more likely than not altered the outcome in
      the case.”
 

Rather, where, as here, a petitioner challenges a death
      sentence, “the question is whether there is a reasonable probability that, absent
      the errors, the sentencer . . . would have concluded that the balance of
      aggravating and mitigating circumstances did not warrant death.”
 Id. at 695.
Putman v. Head, 
268 F.3d 1223
, 1248 (11th Cir. 2001) (alterations and ellipses in

original); see also Ferguson v. Sec’y for Dep’t of Corr., 
580 F.3d 1183
, 1198–99

(11th Cir. 2009) (noting that Strickland asks if a different result is “reasonably

probable,” not if it is “possible” (emphases omitted)). Thus, “[i]n assessing

prejudice, we reweigh the evidence in aggravation against the totality of available

mitigating evidence.” 
Wiggins, 539 U.S. at 534 
(emphasis added). We examine


                                           41
             Case: 17-14570     Date Filed: 07/13/2020    Page: 42 of 56



all of the good and all of the bad, what was presented during the trial and what was

offered collaterally later. The question is whether, “viewed as a whole and

cumulative of mitigation evidence presented originally,” there is “‘a reasonable

probability that the result of the sentencing proceeding would have been different’

if competent counsel had presented and explained the significance of all the

available evidence.” 
Williams, 529 U.S. at 399
. “A reasonable probability is a

probability sufficient to undermine confidence in the outcome” -- in this case, a

probability sufficient to undermine confidence that the jury would have

recommended death. 
Strickland, 466 U.S. at 694
. In determining whether a

reasonable probability of a different outcome exists, we presume a reasonable

decisionmaker. See Nix v. Whiteside, 
475 U.S. 157
, 175 (1986) (“[I]n judging

prejudice and the likelihood of a different outcome, ‘a defendant has no entitlement

to the luck of a lawless decisionmaker.’” (alteration adopted) (quoting 
Strickland,

466 U.S. at 695
)).

      Before addressing the merits of Dallas’s ineffective-assistance-of-counsel

claim, we note a patch of procedural underbrush, though it does not affect our

analysis. The parties agree that petitioner’s claim was procedurally defaulted in

state court, although they disagree sharply about when the default occurred and

thus its effect. Dallas argues the default happened when the state trial court said

that he had “abandoned” the claim by failing to present the testimony of relevant


                                          42
              Case: 17-14570     Date Filed: 07/13/2020    Page: 43 of 56



witnesses at his Rule 32 hearing, but that ineffective assistance of counsel in that

proceeding excuses the default under Martinez v. Ryan, 
566 U.S. 1 
(2012); the

state says the claim was adjudicated on the merits by the state circuit court and not

defaulted until the Alabama Court of Criminal Appeals dismissed it when Dallas

failed to timely appeal the trial court’s denial of his Rule 32 petition.

      Ultimately, when the claim was procedurally defaulted is of no moment

because Dallas’s claim fails on the merits. As we have said many times and as the

Supreme Court has held, a federal court may skip over the procedural default

analysis if a claim would fail on the merits in any event. 
Loggins, 654 F.3d at

1215 
(“When relief is due to be denied even if claims are not procedurally barred,

we can skip over the procedural bar issues, and we have done so in the past.”);
 

see 
Bell, 543 U.S. at 451 
n.3 (citing § 2254(b)(2) for the

proposition that “an application for habeas corpus may be denied on the merits,

notwithstanding a petitioner’s failure to exhaust in state court”); see also Berghuis

v. Thompkins, 
560 U.S. 370
, 390 (2010) (explaining that courts can “deny writs of

habeas corpus under § 2254 by engaging in de novo review when it is unclear

whether AEDPA deference applies, because a habeas petitioner will not be entitled

to a writ of habeas corpus if his or her claim is rejected on de novo review”). That

is what the district court did here, even though by doing so it obligated itself to

examine the ineffectiveness claim de novo and without affording any deference to


                                           43
                Case: 17-14570        Date Filed: 07/13/2020        Page: 44 of 56



the state court’s judgment. We, like the district court, take the “Ockham’s razor”

approach and analyze the claim de novo.4

       We too conclude that this claim must be denied because Dallas cannot

establish prejudice. When reweighing the aggravating circumstances against the

totality of the mitigating evidence -- again, what was introduced at his original trial

and what Dallas presented in his postconviction proceedings -- we consider the

cumulative nature of the evidence. Mitigating evidence in postconviction

proceedings is cumulative “when it tells a more detailed version of the same story




4
  To do otherwise would require us to engage in a complicated analysis that may well wind up at
de novo review anyway. That analysis would go something like this: To determine whether the
claim was procedurally defaulted at the state trial court level, we would ask whether the
alternative “merits” determination by the state trial court was indeed an “adjudication on the
merits” or instead, as Dallas argues, an invalid (under Alabama law) second review of the
sufficiency of the pleadings, after the state trial court had determined the pleadings sufficient to
entitle Dallas to an evidentiary hearing. If the claim was defaulted at the trial court level, we
would then conduct a Martinez analysis to determine whether ineffective assistance of counsel in
his Rule 32 proceeding excuses the default, and if it does, review the claim on the merits. And if
the claim was not defaulted at the trial court level, we would then analyze whether the claim was
instead procedurally defaulted at the appellate level. That analysis is complicated by the
procedural history. Dallas argues that his motion for reconsideration in the trial court should
have tolled the jurisdictional period for filing an appeal under Alabama law -- and he says that
the state appellate court’s refusal to toll the jurisdictional deadline was not a “firmly established
and regularly followed” state procedural rule that would be adequate to bar federal review of his
claim. See Walker v. Martin, 
562 U.S. 307
, 316 (2011) (“To qualify as an ‘adequate’ procedural
ground, a state rule must be ‘firmly established and regularly followed.’” (quoting Beard v.
Kindler, 
558 U.S. 53
, 60–61 (2009))). We need not get into any of this difficulty because the
claim fails even under de novo review -- that is, it fails even if the procedural default occurred at
the trial court level and Martinez excused the default, as Dallas claims. While the state maintains
that the claim was procedurally defaulted at the appellate level, the primary argument offered to
this Court is that the claim fails even under de novo review because Dallas cannot establish
prejudice. Thus, we skip over the difficult procedural default questions and cut to the heart of
the matter: whether Dallas can establish prejudice as a result of the claimed failure of his trial
counsel to conduct and present additional mitigating evidence.
                                                 44
              Case: 17-14570     Date Filed: 07/13/2020    Page: 45 of 56



told at trial or provides more or better examples or amplifies the themes presented

to the jury.” Holsey v. Warden, Ga. Diagnostic Prison, 
694 F.3d 1230
, 1260–61

(11th Cir. 2012). The Supreme Court has found evidence cumulative where it

“substantiate[s],” “support[s],” or “explain[s]” more general testimony provided at

trial. Cullen v. Pinholster, 
563 U.S. 170
, 200–01 (2011) (“There is no reasonable

probability that the additional evidence Pinholster presented in his state habeas

proceedings would have changed the jury’s verdict. The ‘new’ evidence largely

duplicated the mitigation evidence at trial. School and medical records basically

substantiate the testimony of Pinholster’s mother and brother. Declarations from

Pinholster’s siblings support his mother’s testimony that his stepfather was abusive

and explain that Pinholster was beaten with fists, belts, and even wooden boards.”).

      The story presented at Dallas’s trial was a tragic account of childhood

neglect and abuse, violence, poverty, a dearth of any positive or guiding influence,

and pronounced substance abuse from a very early age. Indeed, the entire defense

strategy was to establish that Dallas never intended to kill Mrs. Liveoak; virtually

every piece of evidence presented bore on that theory, and the story of his life and

substance abuse were its pillars. Most of the testimony offered in the 2007

declarations in district court is cumulative of the mitigating evidence presented at

trial -- that is, many of the declarations simply add some details, substantiate, or

explain some aspects of Dallas’s life that had already been graphically presented,


                                          45
              Case: 17-14570     Date Filed: 07/13/2020    Page: 46 of 56



and in detail at the trial. This evidence merely “amplifies the themes [Dallas]

presented to the jury” at trial. See 
Holsey, 694 F.3d at 1260
–61.

      For starters, one major theme found in the 2007 affidavits was the severe

mental illness that plagued Dallas’s mother, Elaine -- issues described in detail by

Jimmy, Cindy, Paul, and Elaine herself. But Cindy testified at trial that living with

their mother was “hell,” and specifically, that the police and an ambulance had

come to their home to take their mother “to an insane asylum a couple of times.”

Although the 2007 affidavits amplify the theme of growing up with a mentally ill

mother, the trial jury was fully aware of Elaine’s mental illness, including the fact

that she had been hospitalized on account of mental illness on several occasions.

      Another theme presented in the 2007 affidavits concerned the impoverished

conditions of Dallas’s childhood, including the condition of their home and the

lack of food and clothing. But Cindy also testified at trial that living in “that

home” was “hell,” that the Dallas children were “lucky [if they] got food

sometimes,” that “[s]ometimes it would be a week before [they] would get to eat,”

and that they sometimes would not attend school because they had no clean clothes

to wear. The jury also learned that Dallas often skipped school and dropped out in

the sixth grade. Again, the 2007 affidavits amplify these themes. Thus, they are

cumulative of the same themes Dallas presented at the guilt and penalty phases of

his trial and were considered by the jury.


                                           46
              Case: 17-14570     Date Filed: 07/13/2020    Page: 47 of 56



      A third theme raised in the 2007 affidavits was the presence of significant

violence in Dallas’s childhood home. Jimmy, Cindy, Paul, and Elaine all testified

about it. Jimmy said, for instance, that their mother threatened to kill the children

with a butcher knife; Cindy testified that their father was a violent man who beat

their mother in front of the children and chased her around the house with knives

and guns and that their mother beat her; Paul testified that their parents were

violent with each other and that their mother, in turn, beat the children; and Elaine

herself testified that as a result of her drinking, she was abusive to her children,

including spanking them with a paddle. But still again, Cindy testified at trial that

there was “[a] lot of violence and problems” in their home, that she had witnessed

her parents chasing each other with knives, and that their mother beat her, all of

which Donald Dallas witnessed as a child. The 2007 affidavits augment this

testimony, but again, in doing so, they simply amplify the theme of violence found

in the home that was presented to the jury.

      The 2007 affidavits also highlighted the lack of any role model in Dallas’s

life. Jimmy explained that he had positive role models growing up -- role models

that Donald lacked. He said he realized later in life, through counseling, that he

and his siblings were never taught right from wrong by their parents, and that while

he had some positive influences and some structure in his life, his siblings did not

and thus, they “really never had a chance in life.” But, at trial, both Cindy and


                                           47
                 Case: 17-14570   Date Filed: 07/13/2020   Page: 48 of 56



Paul testified about the lack of any guidance or positive role model in the

children’s lives. Cindy testified at trial that their parents “didn’t pay much

attention to” Donald and that “[n]obody cared.” While Jimmy did not testify at

trial, Paul explained to the jury that Jimmy went on to lead a successful life

because he went to live with their grandparents, but that, like Donald, Paul had

been in significant legal trouble and suffered from substance abuse. And Donald

himself testified at trial that he received no guidance on right from wrong and was

left to his own devices as a kid. Although the 2007 affidavits may offer a fuller

account about the absence of a role model, this theme was clearly presented to the

jury at trial.

       Perhaps most important, much of the testimony in the 2007 affidavits

addressed Dallas’s substance abuse, including how his personality and priorities

changed after he became addicted. All of this testimony was cumulative. The

defendant’s main trial strategy was to negate the intent element of capital murder

by focusing on Dallas’s substance abuse, beginning at an early age and continuing

through the murder of Hazel Liveoak. Dr. Renfro told the jury that Dallas’s

substance abuse began at a very early age -- alcohol at age seven or eight,

marijuana at twelve or thirteen, and cocaine and the intravenous use of crystal

methamphetamine at thirteen or fourteen. He explained at some length that drug

abuse at so tender an age makes it harder to quit and impairs development. He also


                                           48
             Case: 17-14570     Date Filed: 07/13/2020   Page: 49 of 56



testified about the powerful psychological addictiveness of crack and the

“desperation that builds in when people are using large volumes of cocaine,”

making them “more prone[] to use violence.” The 2007 affidavit testimony from

Dallas’s family members about how his character changed as he began using

drugs, and the expert testimony offered by Dr. Benedict about Dallas’s substance

abuse, plainly was cumulative. Related to substance abuse, the 2007 affidavits also

highlighted the control over Donald exercised by his partner and accomplice -- and

the person with whom he began using crack cocaine -- Polly Yaw. But Yaw’s

influence on Donald was amply presented to the jury at the penalty phase in the

testimony of Cindy Dallas, Pam Cripple, and Rhonda Chavers.

      The long and short of it is, as we’ve said before, that “no prejudice can result

from the exclusion of cumulative evidence,” Ledford v. Warden, Ga. Diagnostic &

Classification Prison, 
818 F.3d 600
, 649–50 (11th Cir. 2016), and Dallas did not

suffer prejudice from the failure to present the cumulative evidence contained in

the 2007 affidavits.

      There are, however, two pieces of “new” evidence found in the

postconviction record: (1) Dr. Benedict’s diagnosis that Dallas suffered from

learning disorders and ADHD; and (2) the allegations made by Paul Dallas that he

and Donald were sexually assaulted on some four occasions.




                                         49
             Case: 17-14570     Date Filed: 07/13/2020    Page: 50 of 56



      First, Dr. Benedict’s testimony would not have added much beyond the

testimony the jury heard from Dr. Renfro. While it is true that Dr. Benedict

diagnosed Dallas with learning disorders and ADHD, the effects of these

conditions were similar to what Dr. Renfro, Dallas himself, and other witnesses

told the jury at trial. Benedict said that people with Dallas’s learning and attention

disorders are more likely to develop substance abuse disorders during early

adolescence. But the jury already learned, and in detail, from Dr. Renfro, from

Dallas, and from his siblings that the defendant suffered from substance abuse and

at a very early age. Benedict also said these disorders often cause an individual to

drop out of school. But again, the jury heard from Dallas that he had difficulty in

school and in fact dropped out in the sixth grade. Most importantly, Benedict

asserted that individuals with Dallas’s combination of learning, attention, and other

mental health issues are often “unassertive” and “passive-dependent.” But still

again the jury heard ample testimony at trial that Dallas was passive and

unassertive, and that he was under the domination and control of Yaw. Quite

simply, Benedict’s testimony would not have changed the characteristics and

difficulties the jury heard and considered before it recommended that Dallas be

sentenced to die, particularly since the jury already knew that Dallas’s substance

abuse and difficulties in school began at so young an age. While it is true the

learning disorders and ADHD diagnosis would have offered another possible


                                          50
             Case: 17-14570     Date Filed: 07/13/2020    Page: 51 of 56



explanation for some of his difficulties, the jury heard a great deal about the

potential causes of those difficulties that were beyond Dallas’s control, including

the incredibly neglectful and abusive parenting he endured.

      Perhaps most critically, Dr. Benedict’s marginally helpful testimony

regarding the ADHD diagnosis would have come at the cost of undermining part of

Dr. Renfro’s helpful trial testimony. Dr. Renfro categorized Dallas as having

below-average intelligence, while Dr. Benedict said Dallas was “of average

intellectual ability,” though he acknowledged that wide variance among Dallas’s

intelligence tests made it difficult to summarize his general intelligence with one

score. Introducing the ADHD diagnosis would have opened the door to Benedict’s

testimony that Dallas was of “average intelligence” -- testimony that would have

been harmful to Dallas since it would have undermined Renfro’s assessment.

Indeed, “both the Supreme Court and this Court have consistently ‘rejected [the]

prejudice argument [ ] where mitigation evidence was a two-edged sword or would

have opened the door to damaging evidence.’” Ponticelli v. Sec’y, Fla. Dep’t of

Corr., 
690 F.3d 1271
, 1296 (11th Cir. 2012) (alterations in original) (quoting

Cummings v. Sec’y for the Dep’t of Corr., 
588 F.3d 1331
, 1367 (11th Cir. 2009)).

      In short, we readily conclude that there is no reasonable probability that

Benedict’s testimony, even when added together with the additional (and largely




                                          51
             Case: 17-14570      Date Filed: 07/13/2020   Page: 52 of 56



cumulative) postconviction testimony of Dallas’s family members, would have

changed Dallas’s sentence.

      The second piece of new evidence is more troubling. In his 2007 affidavit,

Paul Dallas, who testified at trial, alleged for the first time that he and Donald were

both sexually assaulted. But although Paul’s new allegation paints a darker picture

of Dallas’s childhood, it does not standing alone raise a reasonable probability that

the jury would not have recommended that Dallas be sentenced to death. We note

in passing that Donald Dallas gave his attorneys no indication that this abuse

occurred. See Williams v. Head, 
185 F.3d 1223
, 1237 (11th Cir. 1999) (“An

attorney does not render ineffective assistance by failing to discover and develop

evidence of childhood abuse that his client does not mention to him.”). Moreover,

Paul Dallas testified at his brother’s trial. His testimony immediately followed that

of his sister, Cindy, who alleged in her testimony that she had been molested as a

child. When Paul Dallas took the stand, Dallas’s counsel asked if he “ha[d] any

differences” with the characterization of their upbringing and background that his

sister had presented. Despite Cindy having mentioned the sexual abuse she had

suffered and despite Paul being present for that testimony, and only later saying

that he had witnessed the sexual abuse of Donald, Paul Dallas said simply, “No,

sir.” We have no way of knowing if, had Dallas’s attorneys done anything

differently, Paul would have testified about sexual abuse at the trial.


                                          52
              Case: 17-14570     Date Filed: 07/13/2020    Page: 53 of 56



      Ultimately, and most critically, however, the aggravating factors were

overwhelming, and adding the allegation of sexual abuse would not have

sufficiently changed the balance of those factors or given rise to a reasonable

probability that the outcome would have changed. As we have already discussed,

the jury heard many details of the abusive and poverty-stricken conditions of

Dallas’s childhood. But the jury also necessarily found when it convicted him of

capital murder that Donald Dallas placed an elderly woman in the trunk of a car

and intended for her to die there. That is, that he intended for her to suffer a slow

and agonizing death in the sweltering trunk of a car. That he intended for her to

bear the mental anguish of being essentially buried alive. Of waiting in vain for

hope that would never arrive despite his promises to her. That he intended for her

to endure, in the words of Edgar Allen Poe, “a degree of appalling and intolerable

horror from which the most daring imagination must recoil”: the “unendurable

oppression of the lungs . . . -- the clinging to the death garments -- the rigid

embrace of the narrow house -- the blackness of the absolute Night -- the silence

like a sea that overwhelms . . . -- these things, with the thoughts of the air and grass

above, with memory of dear friends who would fly to save us if but informed of

our fate.” Edgar Allen Poe, The Premature Burial (1850). And that he abducted,

robbed, and nearly inflicted the same fate on 80-year-old Wesley Portwood, who

just three days earlier had warned Dallas that it was too hot to get into the trunk of


                                           53
              Case: 17-14570     Date Filed: 07/13/2020    Page: 54 of 56



a car because he would “smother to death in there.” The jury found that Dallas

intended to inflict unimaginable cruelty, and there is no reasonable probability that

the isolated allegation of sexual abuse, given all else that it knew about Dallas’s

tragic life and about his crime, would have changed the jury’s decision.

      In each of the key Supreme Court cases finding prejudice as a result of

counsel’s failure to offer mitigating evidence, the disparity between what was

presented at trial and what was offered collaterally was vast. In other words, the

balance between the aggravating and mitigating evidence at trial and in

postconviction proceedings shifted enormously, so much so as to have profoundly

altered each of the defendants’ sentencing profiles. In Wiggins v. Smith, trial

counsel introduced no evidence whatsoever about Wiggins’s tragic life history,

which the postconviction record demonstrated was marked by “severe privation

and abuse in the first six years of his life while in the custody of his alcoholic,

absentee mother,” followed by “physical torment, sexual molestation, and repeated

rape during his subsequent years in foster 
care.” 539 U.S. at 535
. In Williams v.

Taylor, trial counsel put on almost no mitigation case, calling witnesses who

testified only generally that Williams was a “nice boy” and not violent, while the

postconviction evidence “dramatically described mistreatment, abuse, and neglect

during his early childhood,” and also contained testimony “that he was ‘borderline

mentally retarded,’ had suffered repeated head injuries, and might have mental


                                           54
             Case: 17-14570     Date Filed: 07/13/2020    Page: 55 of 56



impairments organic in 
origin.” 529 U.S. at 369
–70. In Porter v. McCollum, trial

counsel put on nothing in mitigation except “inconsistent testimony about Porter’s

behavior when intoxicated and testimony that Porter had a good relationship with

his son,” while the postconviction record painted a severely abusive childhood,

including routinely witnessing his father beating his mother, as well as being the

target himself of his father’s violence, along with a heroic and decorated record of

military service that left him with post-traumatic stress disorder and brain damage.

558 U.S. 30
, 32–36 (2009). And in Andrus v. Texas, trial counsel “performed

almost no mitigation investigation, overlooking vast tranches of mitigating

evidence,” leading counsel to present a startlingly weak mitigation case that not

only failed to give the jury any insight into the defendant’s tragic childhood, but

ultimately “backfired by bolstering the State’s aggravation case.” No. 18-9674,

slip op. at 9 (S. Ct. June 15, 2020) (per curiam). The postconviction record there

demonstrated that Andrus “suffered ‘very pronounced trauma’ and posttraumatic

stress disorder symptoms from, among other things, ‘severe neglect,’ and exposure

to domestic violence, substance abuse, and death in his childhood,” none of which

his counsel investigated or presented.
 Id. at 10.
      Unlike in Wiggins, in Williams, in Porter, and in Andrus, the new mitigating

evidence contained in the 2007 affidavits “would barely have altered the

sentencing profile presented” at Dallas’s trial. 
Strickland, 466 U.S. at 700
. That


                                          55
             Case: 17-14570     Date Filed: 07/13/2020   Page: 56 of 56



profile amply painted a broad picture of Donald Dallas’s life: an abusive

childhood, violence, poverty, lack of guidance and role models, and substance

abuse from an early age into adulthood. Recognizing that the vast majority of the

allegedly new mitigating evidence presented in the 2007 affidavits did no more

than amplify the themes presented at trial, we think it wholly unlikely that the

additional evidence would have changed the jury’s result. Our confidence that the

jury would have recommended death has not been undermined. In the face of the

horrific nature of Dallas’s crime and the brutality of Hazel Liveoak’s death, and

because the jury already knew much about Dallas’s life, there is no reasonable

probability that, had the jury known the limited additional details presented in

postconviction, they would have spared his life.

      We, therefore, AFFIRM the district court’s denial of Dallas’s § 2254 habeas

petition.




                                         56