[PUBLISH]


             IN THE UNITED STATES COURT OF APPEALS

                    FOR THE ELEVENTH CIRCUIT                      FILED
                      ________________________           U.S. COURT OF APPEALS
                                                           ELEVENTH CIRCUIT
                                                               AUG 23, 2010
                             No. 09-16090
                                                               JOHN LEY
                       ________________________                 CLERK

                   D. C. Docket No. 06-00084-CV-WLS

MARCUS R. JOHNSON,



                                                          Petitioner-Appellant,

                                  versus

STEVEN UPTON,

                                                        Respondent-Appellee.


                       ________________________

                Appeal from the United States District Court
                    for the Middle District of Georgia
                     _________________________

                            (August 23, 2010)


Before CARNES, HULL and PRYOR, Circuit Judges.
HULL, Circuit Judge:

      Marcus Ray Johnson, a Georgia prison inmate under a death sentence,

appeals the district court’s denial of his 28 U.S.C. § 2254 petition for a writ of

habeas corpus. The issues on appeal concern whether Johnson’s trial counsel were

ineffective in the penalty phase as to evidence of Johnson’s life history, escape

from pretrial custody, and future dangerousness. After review and oral argument,

we conclude the Georgia state court’s denial of Johnson’s ineffective counsel

claims was not based on an unreasonable determination of the facts or an

unreasonable application of clearly established federal law. Thus, we affirm.

                         I. PRETRIAL BACKGROUND

A.    The Crime

      On March 24, 1994, Johnson raped, murdered, and mutilated Angela

Sizemore a few hours after meeting her at a bar in Albany, Georgia. See Johnson

v. State, 

, 225 (Ga. 1999). The Georgia Supreme Court set forth the

evidence against Johnson:

      [T]he victim, Angela Sizemore, met Johnson in a west Albany bar
      called Fundamentals between 12:30 and 1:30 a.m. on March 24, 1994.
      Ms. Sizemore had been to a memorial service for an acquaintance the
      previous day, and she had been drinking so heavily the bar had
      stopped serving her. Johnson was wearing a black leather jacket,
      jeans, black biker boots, and a distinctive turquoise ring. According

                                           2
to a witness, Johnson was angry and frustrated because another
woman had spurned his advances earlier in the evening. The bar
owner and its security officer (who both personally knew Johnson)
testified that they saw Johnson and Ms. Sizemore kissing and
behaving amorously. [At approximately 2:30 a.m.] Johnson and Ms.
Sizemore left Fundamentals together; the bartender handed Ms.
Sizemore’s car keys directly to Johnson. They were seen walking
towards Sixteenth Avenue.
        At approximately 8:00 a.m. on March 24, 1994, a man walking
his dog found Ms. Sizemore’s white Suburban parked behind an
apartment complex in east Albany, on the other side of town from
Fundamentals. Ms. Sizemore’s body was lying across the front
passenger seat. . . .
        Four people testified that they saw Johnson about an hour
before the body was found. Two witnesses testified that [at around
7:00 a.m.] they saw him walk from the area where the victim’s
Suburban was parked through an apartment complex to a bus stop.
He boarded a bus and asked if the bus would take him to the Monkey
Palace (a bar where Johnson worked) in west Albany. Three
witnesses, including the bus driver, identified Johnson as being on the
bus (one of the witnesses who saw Johnson walk through the
apartment complex boarded the same bus as he did). Two witnesses
stated that their attention was drawn to Johnson because that area of
Albany is predominantly African-American, and it was extremely
unusual to see a Caucasian there at that time of day. All the witnesses
testified that Johnson’s clothes were soiled with dirt or a substance
they had assumed to be red clay. The witnesses gave similar
descriptions of his clothing; in court, two witnesses who sat near
Johnson on the bus identified his jacket, boots and distinctive
turquoise ring.
        The police determined that Ms. Sizemore was murdered in a
vacant lot near Sixteenth Avenue in west Albany. . . . The vacant lot
is about two blocks from Fundamentals and about half a block from
the house where Johnson lived with his mother.
        A friend of Johnson testified that after he called her early on
March 24, she picked him up at his house at 9:30 a.m. and took him to
her home, where he slept on her couch for several hours. Johnson
then told her he wanted to take a bus to Tennessee and that he needed

                                   3
       her to go to the Monkey Palace to pick up some money he was owed.
       At his request, she dropped him off near a church while she went to
       get the money. The police were waiting for Johnson to show up, and
       they returned with the friend and arrested Johnson. Before they told
       him why they were arresting him, he blurted, “I’m Marcus Ray
       Johnson. I’m the person you’re looking for.”
              DNA testing revealed the presence of the victim’s blood on
       Johnson’s leather jacket.        Johnson had a pocketknife that was
       consistent with the knife wounds on the victim’s body. He had
       scratches on his hands, arms, and neck. In a statement, Johnson said
       he and the victim had sex in the vacant lot and he “kind of lost it.”
       According to Johnson, the victim became angry because he did not
       want to “snuggle” after sex and he punched her in the face. He stated
       he “hit her hard” and then walked away, and he does not remember
       anything else until he woke up after daybreak in his front yard. He
       said, “I didn’t kill her intentionally if I did kill her.”

       The condition of Sizemore’s body evidenced Johnson’s extreme brutality

during her murder. Johnson sexually assaulted Sizemore with the limb of a pecan

tree, which was shoved into her vagina until it tore through the back wall of her

vagina and into her rectum. Sizemore was alive during the sexual assault.

       Johnson also cut and stabbed Sizemore 41 times with a small, dull knife.

Sizemore had grip marks (round or oval bruises caused when a person is grabbed

tightly) on her upper extremities, knees, thighs, ankles, and the inside of her arms.

She had severe bruising, abrasions, and other evidence of blunt trauma about her

body, especially her face, head, arms, ankles, and feet. Sizemore was alive during

this attack.

                                           4
      After mutilating and killing Sizemore, Johnson dragged her body from the

attack area back to her car. Sizemore’s body was discovered clothed, with her shirt

pulled up and tied in a knot just below the breast area. Her pants were around her

legs and her bra was tied in a knot around her right thigh and protruded from the

pants. Dirt and sand drag marks were found on the side of her body and grass was

found attached to her face. Johnson had dragged Sizemore’s body from the attack

area back to her car by using the knotted loops of her shirt and bra as handles.

B.    Appointment of Counsel

      The state trial court provided Johnson with two exceptionally well-qualified

criminal defense attorneys. Four days after Johnson’s arrest, the state trial court

appointed experienced criminal trial attorney Ronnie Joe Lane.1 Lane had

practiced criminal law for almost 20 years and handled hundreds of criminal cases,

including about 40 murder trials. In all four of his previous capital cases, Lane

secured his clients life sentences. Lane tried two death penalty cases to life

sentences and pled two other capital cases to life sentences.

      On August 15, 1994, the State announced its intent to seek the death penalty.

Johnson did not go to trial until almost four years later, in March 1998. In June

1997, at Lane’s request, the state trial court appointed attorney Tony Jones to assist



      1
          Lane currently serves as a Superior Court judge for the Pataula Judicial Circuit.

                                                  5
Lane.2 Jones had practiced criminal law for 14 years and had handled numerous

felony cases, including at least two murder cases. Lane and Jones served as co-

counsel at trial.

C.     Johnson’s Transfer to Miller County Jail

       Following his arrest, Johnson was housed in the overcrowded Dougherty

County jail in Albany, Georgia.3 On June 5, 1994, Johnson was taken to a hospital

for treatment after other inmates beat him.4 Lane knew that Johnson received other

beatings from jail inmates.5 Lane saw Johnson had suffered injuries, but they were

“mainly bruises and lacerations, . . . not anything that would have required him to

be hospitalized.” Johnson gave Lane handwritten threatening notes that Johnson

received in the Dougherty County jail. Lane “knew it wasn’t a healthy situation

for [Johnson] to be there.”


       2
        Lane waited to request co-counsel until he determined he would not challenge venue
because he wanted a local attorney as co-counsel.
       3
          In January 1995, Dougherty County opened a second, newer jail. We refer to that
facility as the “new Dougherty County jail” and the first jail as either the “old Dougherty County
jail” or simply the “Dougherty County jail.”
       4
         Records show Johnson was transported to the emergency room and diagnosed with
contusions to his neck, arms, and face. X-rays revealed no fractures. Johnson complained of pain
in his left wrist and left eye, suffered swelling in his left wrist, and had a knot on his forehead.
       5
        Lane did not know of any alleged rapes. Johnson in fact denied being raped or sexually
assaulted. In July 1995, Johnson underwent a psychological assessment at Central State
Hospital, as part of which Johnson completed a questionnaire that included the question, “Have
you ever been sexually or physically abused? Explain.” Johnson wrote “No” next to “sexually,”
and next to “physically” he wrote, “physically abused in Dougherty Co. jail.”

                                                 6
      Attorney Lane met with Johnson at the jail on June 6, 1994, to discuss the

jail conditions and his beating. Lane later met with the sheriffs of Dougherty

County and nearby Miller County to discuss transferring Johnson to the Miller

County jail. Lane told the Dougherty County sheriff about Johnson’s abuse in the

Dougherty County jail. Lane also preferred that Johnson be housed in the Miller

County jail because it was closer to Lane’s office in Donalsonville, Georgia.

      After four months in the Dougherty County jail and because of Lane’s

request, Johnson was transferred to the Miller County jail in August 1994.

D.    Johnson’s Escape

      The Miller County jail was a small facility, and on the evening of October 2,

1994, the only deputy on duty was 76-year-old Brooks Sheffield. That evening,

Johnson asked Deputy Sheffield if he could use the jail telephone. When Sheffield

brought Johnson to the telephone, Johnson grabbed Sheffield’s gun, struck

Sheffield in the head with the butt of the gun, and escaped the jail. The next night,

Johnson was found and taken back into custody.

      Sheffield’s head injury required 21 staples to close, plus follow-up care. X-

rays showed no skull fracture or intracranial bleeding. There were no brain

contusions. The CT scan “was deemed normal for a patient in Mr. Sheffield’s age

range.” About seven months later, on April 27, 1995, Sheffield suffered a stroke.



                                          7
He died in June 1995.

      Upon Johnson’s recapture, he was returned to the Miller County jail, where

he remained until mid-November 1994. Johnson was then transferred back to a jail

annex in Dougherty County, which was a separate, renovated building across the

street from the old jail’s cell blocks where Johnson was before. In January 1995,

the new Dougherty County jail opened and Johnson was moved there. Johnson

remained in the new Dougherty County jail until his trial ended.

      According to Lane’s billing records, on October 24, 1994 he met with

Johnson at the Miller County jail and “discussed case, reviewed impact of escape

and need for no further violations.” This meeting, together with thirty miles’

travel, took 3.2 hours. Lane later discussed the escape with an assistant district

attorney. It is undisputed that Johnson never told Lane he escaped because he

feared going back to the old Dougherty County jail.

E.    Trial Counsel’s Penalty-Phase Preparation

      Lane first met with Johnson on the day Lane was appointed, March 28,

1994. Lane had good contact with Johnson and “visited him fairly regularly in

jail.” Johnson called Lane at Lane’s home on average once a week throughout the

four-year representation.

      During the first approximately three years after his appointment, Lane chose



                                           8
to focus his efforts on: (1) challenging Johnson’s tape-recorded statement to police,

which Johnson claimed, and Lane believed, was edited by police officers to erase

Johnson’s request for an attorney; (2) trying to suppress eyewitness identifications

of Johnson and certain items of physical evidence; (3) “interact[ing] with various

experts who were assisting [Lane] with these matters” and other experts who

assisted Lane “on forensic matters related to Ms. Sizemore’s death and the crime

scenes”; and (4) “preparing and litigating numerous pre-trial motions.”6

       Johnson “maintained his innocence throughout” the case. Lane felt Johnson

had a chance to be acquitted because of “the circumstantial nature of the evidence

against Mr. Johnson and the lack of conclusive physical evidence tying him to a

homicide.” Lane testified:

       Compared to other death penalty cases where the evidence is just
       overwhelming that the person did it, this was not such a case. You
       would try that case totally differently. You’d forget about did he do
       it. And you start trying it about, well, why did he do it. This case was
       not like that.

       As part of his pretrial preparation, Lane consulted with other criminal

defense lawyers and with experienced capital attorneys at the Southern Center for

Human Rights and the Multicounty Public Defender’s Office, spoke with a



       6
        After Jones was appointed, he and Lane worked together on the whole case, but Lane
was lead counsel and did most of the work. Jones testified that “Lane was more in charge” of
the penalty phase of the trial than Jones was.

                                               9
mitigation specialist, and attended several death penalty seminars that stressed the

importance of mitigation and the defendant’s social history.

      Lane began his penalty-phase investigation in earnest in 1997, about a year

before trial, although he may have had discussions with Johnson in 1994 or 1995

about “what had happened to [Johnson] in his life.” As detailed later, Lane

discussed Johnson’s background and marital, social, employment, and medical

history not only with Johnson but also with his parents, brother, former girlfriend,

and others. Two investigators were assigned to assist Lane.

      Lane knew the State would use evidence of Johnson’s escape in sentencing.

Lane believed that “the escape incident could prove to be devastating to Mr.

Johnson’s case if [he] proceeded to a sentencing phase” because, in Lane’s

experience, “future dangerousness is of great concern to juries in capital cases, and

an escape clearly raises the specter of future dangerous behavior in the jury’s

eyes.” Lane felt the evidence of Johnson’s escape “would be some of the most

damaging evidence presented.”

      As noted earlier, Lane discussed the escape with Johnson for almost three

hours. Although Johnson now alleges he escaped due to fear of going back to the

old Dougherty County jail, Johnson never told Lane this. Lane acknowledged that

he never asked Johnson specifically what made him escape. Nonetheless, it is



                                          10
undisputed that Johnson never told Lane that he escaped because of his fear of

going back to the old Dougherty County jail. In fact, even after the escape,

Johnson was taken back to the Miller County jail.

          II. TRIAL: STATE’S WITNESSES AT PENALTY PHASE

      From March 23 to April 7, 1998, Johnson was tried on charges of malice

murder, felony murder, aggravated assault, aggravated battery, and rape. The jury

found Johnson guilty of all charges.

      At the penalty phase, the State called eleven witnesses. Lynwood Houston,

a sheriff’s deputy in Miller County, investigated Johnson’s escape and testified

about it. The jailer on duty, Sheffield, allowed Johnson to come into the control

room to use the telephone. Once there, Johnson snatched the pistol out of

Sheffield’s holster and hit Sheffield in the head with it. Johnson fled the jail,

taking Sheffield’s pistol with him.

      On cross-examination, Deputy Houston admitted that Johnson committed no

crimes after his escape that Houston knew of, and that Johnson did not try to run or

resist arrest when he was recaptured. Deputy Houston did not know how long

Brooks Sheffield was in the hospital after Johnson hit him with the gun, but

Sheffield did return to work for the county after the assault. Deputy Houston knew

that after his injury, Sheffield did some work for the city too, and may have



                                           11
directed the public works efforts around the courthouse square.

      Dr. Peggy Rummel, the emergency room physician who treated Sheffield,

testified that Sheffield “came to the emergency room with his head bleeding, and

he stated at the time that he had been struck on his head with a gun butt during an

assault by an escaping prisoner.” Sheffield was 76 years old but was in good

health except for his head injury. Dr. Rummel examined Sheffield, closed the

wound, and sent Sheffield to get a CT scan “to make sure that no brain damage had

resulted from the blow to his head.” The injury came from a “pretty significant

blow” that “caught the skin and had just literally scalped and peeled it down to

where he had exposed bone and this was bleeding very heavily.” It required 21

staples to close. Dr. Rummel opined that the injury resulted from a glancing blow

and, with the amount of force that was used, a direct blow “would have crushed

[Sheffield’s] skull.”

      Dr. Rummel, who knew Sheffield personally, testified that after this

incident, Sheffield “was just not the same man any more.” Sheffield was “a very

unhappy man afterwards” and “stayed home and stayed to himself.” Seven months

after Johnson’s escape, Sheffield “suffered a stroke on that side and it was a bleed

of one of the major arteries in that area and it had disastrous consequences for

him.” Sheffield died several weeks later. Dr. Rummel testified, though, that she



                                          12
“[could ]not in all honesty tell you that there was a one hundred percent no doubt

about it direct relationship” between the head injury and Sheffield’s stroke seven

months later.7 On cross-examination, Dr. Rummel admitted that Sheffield drove

himself to the hospital after receiving the head injury and that the CT scan taken

after the assault revealed no internal bleeding in Sheffield’s brain.

       The State called Johnson’s former probation officers, who testified Johnson

received probation in September 1984 after pleading guilty to financial transaction

card fraud, four counts of financial transaction card theft, and theft by receiving

stolen property.8 One officer testified that of the thousand persons he had

supervised as a parole officer, Johnson’s attitude was one of the ten worst.

Johnson was “[v]ery resistant to supervision” and was rated a “maximum risk

individual.” Another officer confirmed Johnson had a negative attitude and failed

to comply with the terms of his probation.



       7
         Similarly, the State’s penalty-phase closing argument suggested that the injury Johnson
inflicted may have caused Sheffield’s death, but admitted it may not have and maybe Sheffield
would have died anyway:
        [Johnson] slugged a seventy-six year old deputy in the head, so severely that if it had
        been a direct blow it would have killed him. What did happen to him? Several
        months later, he suffered a stroke on the same side of his body that he got that hit;
        his health just kept going straight on down and he died and Mr. Lane or Mr. Jones
        may say, well, he didn’t kill him, and maybe he didn’t and maybe he would have
        died anyway, but that happened at the hands of this man, Marcus Ray Johnson, and
        there’s no dispute of that, doing anything he could to get out of lawful confinement.
(Emphasis added.)
       8
           The State admitted certified copies of Johnson’s convictions, plea, and sentence.

                                                  13
      The State closed its penalty-phase case with powerful victim impact

testimony from five witnesses: Sizemore’s then-eight-year-old daughter, mother,

two sisters, and stepfather. The State also showed the jury a video of the scene

where Sizemore’s body was found.

      Sizemore’s daughter described getting out of bed one morning when she was

four years old and realizing her mother wasn’t there and would never come back.

She testified she still cried in school, and she wanted the man who killed her

mother punished for taking her mother away from her.

      Sizemore’s mother and stepfather described Sizemore’s kindness,

generosity, adventurous spirit, and passion for being a mother. Sizemore’s mother

testified that Sizemore’s death “left a hole in our hearts that time won’t heal.” She

described her emotional struggle in raising her eight-year-old granddaughter while

wondering if she will “learn to hate because of this” or “be distrusting or holding

emotions of this trauma inside to the point that it destroys her ability to be a loving

person.”

      Sizemore’s sisters testified that Sizemore was their role model and best

friend. Sizemore was fearless and passionate and the most dedicated mother they

knew. Sizemore was full of life and did not want to die. One sister testified about

how heartbreaking it was to see the wistful look on Sizemore’s daughter’s face



                                           14
when she looked at other families. This sister wished it was she who had died

instead of her sister because “it would have been easier than the daily hell I go

through thinking about the last hours of my sister’s life.”

           III. TRIAL: DEFENSE’S WITNESSES AND CLOSING
                    ARGUMENT AT PENALTY PHASE

      The defense called four witnesses. The defense’s plan was primarily to

argue residual doubt (also called lingering doubt). Given there were no direct

eyewitnesses to the murder, Lane believed lingering doubt would be paramount in

the penalty phase. He stated, “Lingering doubt was the whole thing in this case.

Without it, I didn’t see any hope.”

      Reverend Bobby Moye was the pastor of Johnson’s church and a friend of

Johnson’s family. Rev. Moye testified that he had known Johnson since childhood

and had baptized Johnson, and that Johnson was a “fine boy” who was “very kind”

and “very involved in church.” On cross-examination, Rev. Moye stated that

Johnson came to church as an adult, but Rev. Moye “could not accurately say how

faithful he was or how dependable.”

      Reverend Jimmy O. Hall, Jr., Johnson’s middle-school physical education

teacher, testified that Johnson was “one of the students that stood out” and was

“very dutiful” and “a very good student as far as participation.” Johnson was “a

pretty good student,” was “always cooperative,” and did not cause trouble.

                                          15
Johnson “would always worry . . . about could I take the equipment out and that

kind of thing. May I lead exercise, things of that nature, may I be captain of the

team, things of that nature.” Rev. Hall sought out defense counsel and offered to

testify on Johnson’s behalf “[b]ecause I knew Ray Johnson . . . because I feel he’s

a good person.” Rev. Hall admitted he did not know Johnson as an adult.

      Brian Sherman, a friend of Johnson’s, testified about his common interests

with Johnson: (1) Johnson was a cabinetmaker and Sherman appreciated Johnson’s

artistry; and (2) Johnson and Sherman both enjoyed dancing. Sherman asked the

jury not to sentence Johnson to death.

      Johnson’s mother, Rosemary Johnson, begged the jury “not [to] put [her]

son to death.” Rosemary Johnson testified that her son’s arrest and trial had

“destroyed [her] whole life . . . . I’ve had to carry the burden right by myself . . .

and I have been right here with my child, and I will be with my child until my

dying day. . . . Or to his dying day.”

      Lane spent most of his penalty-phase closing argument in an attempt to

create lingering doubt. He told the jurors that although they decided Johnson was

guilty, they “should not vote to kill Ray Johnson . . . because this is a

circumstantial evidence death penalty murder case only.” Lane argued the jurors

“should have some doubt, some doubt, whatever, ever so slight, whatever doubt,



                                           16
some doubt.” Lane posed a number of hypothetical questions he argued that the

State had not satisfactorily answered, and asked the jurors whether they were sure

enough in their verdict, “so sure, sure enough based on this circumstantial

evidence, to vote to take a human being’s life?” Lane emphasized that the death

penalty is irreversible.

       Afterward, Lane argued that the State “would have you believe that

[Sheffield] died as a result of this wound that he got on his head” but “[t]hey have

no proof of that.” Lane pointed out that Sheffield had returned to work for the city

and the county after the attack. Lane argued none of the State’s other penalty-

phase evidence was sufficient to justify a death sentence.

       In Jones’s portion of the closing argument, he discussed Johnson’s life,

emphasizing Johnson’s helpfulness as a child and his church attendance. Jones

also argued that Dr. Rummel would not have seen Sheffield much after he received

his head injury because Sheffield was working at the hospital through his capacity

as a Miller County deputy, and he was “no longer in that capacity with the county”

after Johnson’s escape. Johnson argued that the State’s suggestion that Johnson

killed Sheffield was “a most tragic play on the facts” and “unfair.”

       Jones also plied a residual-doubt argument, telling the jury that the State’s

case was circumstantial and that although the jury found Johnson guilty beyond a



                                           17
reasonable doubt, “I’m begging you, before we execute this man, let us take that to

a little higher level.” Jones asked the jury to consider the “unanswered questions

during the course of the trial.”

                      IV. SENTENCE AND DIRECT APPEAL

       The jury returned a unanimous verdict fixing Johnson’s sentence at death.

The jury found beyond a reasonable doubt the existence of four statutory

aggravating circumstances: (1) Johnson murdered Sizemore while he was engaged

in the commission of another capital felony (i.e., rape); (2) Johnson murdered

Sizemore while he was engaged in the commission of aggravated battery; (3)

Johnson’s murder of Sizemore was outrageously and wantonly vile, horrible, and

inhumane in that it involved torture and depravity of mind; and (4) Johnson’s

murder of Sizemore was outrageously and wantonly vile, horrible, and inhumane

in that it involved an aggravated battery. The state trial court sentenced Johnson to

death on the malice murder conviction in accordance with the jury’s

recommendation. The state trial court sentenced Johnson to life imprisonment for

the rape and to 20 years’ imprisonment for the aggravated battery.9


       9
         The aggravated assault conviction merged into the felony murder conviction, and the
felony murder conviction was vacated by operation of law. See 

n.1
(citing Malcolm v. State, 

, 482 (Ga. 1993) (noting that: (1) aggravated assault
conviction, as underlying felony in felony murder count, merged into felony murder conviction
as a matter of law; and (2) a defendant may be convicted in Georgia of both malice murder and
felony murder of single victim, but may not be sentenced for both crimes)).

                                              18
      The Georgia Supreme Court affirmed Johnson’s convictions and sentences.

. The United States Supreme Court denied Johnson’s

petition for certiorari. Johnson v. Georgia, 

, 

(2000).

                      V. STATE HABEAS PROCEEDINGS

      After his direct appeal ended, Johnson obtained new counsel (who still

represents Johnson) and filed a state habeas petition. Johnson’s state petition

claimed, among other things, that he received ineffective trial counsel because

counsel did not adequately investigate and present: (1) evidence of the

circumstances surrounding Johnson’s escape from the Miller County jail; (2) the

cause of Deputy Sheffield’s death; (3) expert testimony and statistical evidence on

Johnson’s likelihood of future dangerousness; and (4) mitigating evidence from

Johnson’s life history.

      The state habeas court held a three-day evidentiary hearing on June 24-26,

2002. Johnson’s new counsel presented testimony from five witnesses, affidavits

from thirty-four additional witnesses, and documentary exhibits. The State called

six witnesses and also introduced documents.

      The state habeas court issued a detailed 36-page order denying all of

Johnson’s claims. The state habeas court stated, “The court finds that the

petitioner has failed to establish deficiency and resulting prejudice as to any of his



                                           19
ineffective assistance of counsel claims.” We discuss the relevant findings of the

state habeas court in our later analysis of Johnson’s claims in this appeal.

      The Georgia Supreme Court denied Johnson’s application for a certificate of

probable cause to appeal the denial of his state habeas petition. The United States

Supreme Court denied Johnson’s certiorari petition. Johnson v. Terry, 

, 

, reh’g denied, 

, 

(2006).

                   VI. FEDERAL HABEAS PROCEEDINGS

      On June 7, 2006, Johnson filed a 28 U.S.C. § 2254 petition in federal district

court. On September 30, 2009, the district court denied Johnson’s petition.

      The district court granted Johnson a COA on three ineffective trial counsel

claims as to the penalty phase: (1) whether “counsel was ineffective for failing to

mitigate the circumstances around [Johnson’s] escape from the Miller County

Jail”; (2) whether “counsel was ineffective for failing to rebut or preclude

testimony regarding the death of jailer Brooks Sheffield”; and (3) whether “counsel

was ineffective in failing to rebut the state’s future dangerous[ness] argument by

using statistical and expert testimony.” This Court expanded the COA to include a

fourth claim: “[w]hether [Johnson’s] trial counsel was ineffective in failing to

investigate or present ‘life history mitigation testimony’ in the penalty phase.”

                         VII. STANDARD OF REVIEW



                                          20
      Section 2254, as amended by the Antiterrorism and Effective Death Penalty

Act of 1996 (“AEDPA”), “constrains our review of legal questions decided on the

merits in state court.” Mason v. Allen, 

, 1119 (11th Cir. 2010).

      Under AEDPA, a federal court may not grant a habeas corpus
      application “with respect to any claim that was adjudicated on the
      merits in State court proceedings,” 28 U.S.C. § 2254(d), 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,” § 2254(d)(1), or “was based on
      an unreasonable determination of the facts in light of the evidence
      presented in the State court proceeding,” § 2254(d)(2).

Berghuis v. Thompkins, 560 U.S. —, 

, 2259 (2010). The Supreme

Court has described this standard as “a highly deferential” one that “demands that

state-court decisions be given the benefit of the doubt.” Renico v. Lett, 559 U.S.

—, 

, 1862 (2010). However, where the petitioner makes the

required § 2254(d) showing as to a state court decision, we owe no AEDPA

deference to that decision and instead review the claim de novo. Jones v. Walker,

, 1288 & n.5 (11th Cir. 2008) (en banc), cert. denied, 

(2009); McGahee v. Ala. Dep’t of Corr., 

, 1266 & n.20 (11th

Cir. 2009).

      “We review de novo the district court’s decision about whether the state

court acted contrary to clearly established federal law, unreasonably applied

federal law, or made an unreasonable determination of fact.” Reed v. Sec’y, Fla.

                                         21
Dep’t of Corr., 

, 1239 (11th Cir. 2010), petition for cert. filed (U.S.

Jun. 7, 2010) (No. 09-11314). Because Johnson raised the ineffective trial counsel

claims at issue here in his state habeas petition, and the Georgia Supreme Court

denied Johnson’s application for a certificate of probable cause to appeal the denial

of that petition, the relevant state court merits decision in this case is the state

habeas court’s order. See 

(stating, where state

supreme court denied discretionary review, that “[t]he relevant state-court decision

here is the Michigan Court of Appeals’ decision . . . rejecting [Petitioner’s]

Miranda and ineffective-assistance-of-counsel claims on the merits”).

                                   VIII. ANALYSIS

       All of Johnson’s claims on appeal assert ineffective assistance of his trial

counsel. Ineffective counsel claims are governed by the two-pronged test

enunciated in Strickland v. Washington, 

, 

(1984):

       A convicted defendant’s claim that counsel’s assistance was so
       defective as to require reversal of a conviction or death sentence has
       two components. First, the defendant must show that counsel’s
       performance was deficient. . . . Second, the defendant must show that
       the deficient performance prejudiced the defense. . . . Unless a
       defendant makes both showings, it cannot be said that the conviction
       or death sentence resulted from a breakdown in the adversary process
       that renders the result unreliable.

104 S. Ct. at 2064.

       For the performance prong, the governing standard is objectively reasonable

                                            22
attorney conduct under prevailing professional norms:

      As to counsel’s performance, “the Federal Constitution imposes one
      general requirement: that counsel make objectively reasonable
      choices.” Bobby v. Van Hook, 558 U.S. –, 

, 17, 

(2009) (quotation marks omitted). Thus, to establish
      deficient performance, a defendant must show that his counsel’s
      conduct fell “‘below an objective standard of reasonableness’ in light
      of ‘prevailing professional norms’” at the time the representation took
      place. 

(quoting 

, 104 S. Ct. at
      2064-65). In assessing the reasonableness of counsel’s performance,
      courts must “indulge a strong presumption that counsel’s conduct falls
      within the wide range of reasonable professional assistance; that is,
      the defendant must overcome the presumption that, under the
      circumstances, the challenged action might be considered sound trial
      strategy.” 

, 104 S. Ct. at 2065 (quotation
      marks omitted). “[S]trategic choices made after thorough investigation
      of law and facts relevant to plausible options are virtually
      unchallengeable; and strategic choices made after less than complete
      investigation are reasonable precisely to the extent that reasonable
      professional judgments support the limitations on investigation.” 

104 S. Ct. at 2066.

.

      For the prejudice prong, the test is whether “there is a reasonable probability

that, but for counsel’s unprofessional errors, the result of the proceeding would

have been different.” 

, 104 S. Ct. at 2068. “To assess

that probability, we consider the totality of the available mitigation evidence – both

that adduced at trial, and the evidence adduced in the habeas proceeding – and

reweigh it against the evidence in aggravation.” Porter v. McCollum, 558 U.S. —,

, 453-54 (2009) (quotation marks and brackets omitted); see Sears v.

                                          23
Upton, — U.S. —, 

, 3267 (2010) (“A proper analysis of prejudice

under Strickland would have taken into account the newly uncovered evidence

. . . , along with the mitigation evidence introduced during Sears’ penalty phase

trial, to assess whether there is a reasonable probability that Sears would have

received a different sentence after a constitutionally sufficient mitigation

investigation.”).

A.    Trial Counsel Performance: Evidence of Johnson’s Escape

      Johnson claims Lane unreasonably failed to investigate the circumstances of,

and reasons for, Johnson’s escape from the Miller County jail. Johnson contends a

reasonable investigation would have revealed extenuating circumstances (Johnson

feared a return to the Dougherty County jail, where he was repeatedly beaten and

abused) that Lane should have presented to rebut the State’s argument that the

escape and attack on Sheffield showed Johnson would be a danger in prison.

      The state habeas court concluded that counsel’s performance was not

deficient. The state habeas court expressly found that Johnson did not tell Lane

about this now-alleged motivation to escape, so Lane had no reason to investigate

further:

      The Petitioner contends that counsel failed to mitigate the damage that
      petitioner’s escape prior to trial [caused] by putting forward the claim
      that petitioner was in some way justified in this escape because of his
      fear of the Dougherty County Jail. During the hearing, counsel for the

                                           24
      respondent asked Mr. Lane (now Judge Lane), “Did he ever tell you
      he escaped from the jail because he was afraid he was going to have to
      go back to the Dougherty County Jail?” The answer given was, “No, I
      never asked him.”
      ....
             The court also finds that the evidence in this case shows that
      prior to trial, Petitioner never told trial counsel that he assaulted the
      guard and escaped because of his alleged fear of being returned to the
      “old” Dougherty County Jail. . . .
      ....
             An attorney does not render ineffective assistance by not
      discovering mitigating evidence that his client did not mention to him.

This decision is not contrary to, or based on an unreasonable application of, clearly

established Supreme Court precedent. Nor is it based on an unreasonable

determination of fact.

      In Strickland, the Supreme Court explained that “[t]he reasonableness of

counsel’s actions may be determined or substantially influenced by the defendant’s

own statements or actions” because “[c]ounsel’s actions are usually based, quite

properly on . . . information supplied by the 

, 104 S.

Ct. at 2066. Thus, “inquiry into counsel’s conversations with the defendant may

be critical to a proper assessment of counsel’s investigation decisions.” Id.; see

Peterka v. McNeil, 

, 1208-09 (11th Cir. 2008) (concluding counsel’s

performance was not deficient for not learning of evidence that petitioner declined

an opportunity to escape with his cellmates between penalty phase and trial court’s

sentencing because, even if counsel had duty to explore further mitigation at this

                                          25
point, petitioner should have known this was evidence that counsel would be

interested in as mitigation but petitioner did not tell counsel about it).

       Here, it is undisputed that after Johnson’s escape and recapture, Lane met

with Johnson in person for approximately three hours, at which time Lane

“reviewed [the] impact of [Johnson’s] escape and [the] need for no further

violations.” It is also undisputed that Johnson did not – either at this three-hour

meeting or at any other time during the four years before trial – tell Lane that he

escaped from the Miller County jail because he feared being returned to the

Dougherty County jail. Thus, Johnson did not tell Lane the now-purported reason

for his escape despite having an extensive discussion about the impact that the

escape would have at trial. Given the charges against him, Johnson was facing a

potential death sentence and, at a minimum, life imprisonment. It is not

objectively unreasonable for an attorney to assume that if there were some

powerful reason for his client’s escape from pretrial detention other than the usual

reason for escapes (not wanting to be convicted and in jail), the client would have

told him that reason when they discussed the impact the escape evidence would

have at trial.

       Nor did Lane have any reasonable basis for believing further investigation

into the unrevealed reason for Johnson’s escape would be likely to yield fruitful



                                           26
information. Although Lane knew that Johnson had been beaten by other inmates

at the Dougherty County jail, Johnson escaped from the Miller County jail, where

he was treated well. And there was no actual transfer of Johnson in the offing that

Lane would have or could have known about.10 Moreover, after his escape

Johnson was taken back to the Miller County jail.11 Lane cannot be deficient for

failing to investigate whether Johnson escaped because he was afraid of a non-

existent jail transfer when “there is no indication that [Johnson] . . . gave [Lane]

reason to believe that such evidence might exist.” Lambrix v. Singletary, 

, 1505 (11th Cir. 1996).

       Johnson argues the state habeas court’s conclusion that Lane did not perform

unreasonably is contrary to Rompilla v. Beard, 

, 

(2005), in which the Supreme Court concluded petitioner’s trial counsel were

deficient in failing to review the court file on his prior conviction that counsel



       10
         Johnson himself did not testify in the state habeas proceeding. The most Johnson
presented was the testimony of Ronnie Daniels (a/k/a “Sneaky D”), a former Miller County
inmate, who testified that Johnson escaped after overhearing a conversation about another
inmate being transferred to Dougherty County and mistakenly assuming the conversation was
about him. Another fellow Miller County jail inmate, James Wright, testified that he spoke with
Johnson after the escape and Johnson “told me that he was going to be transferred back to
Dougherty County. He just kept saying, ‘I can’t go back there, I can’t go back.’”
       11
          Johnson states that one of Lane’s investigators, George Waldrop, “somehow learned
prior to trial” that Johnson escaped from the Miller County jail because he was under the
mistaken belief that he was about to be transferred back to the Dougherty County jail. This is
not correct. In his affidavit, Waldrop expressly states, “I didn’t know anything about [Johnson’s
escape] until it was presented at sentencing.”

                                               27
knew the State intended to use at trial. 

125 S. Ct. at 2464. But

Rompilla is materially different because it concerned trial counsel’s failure to

review a readily available public document that the State announced at least twice

that it would rely upon at trial (i.e., the court file on the defendant’s prior

conviction, which included (1) psychological test results pointing to schizophrenia

and other disorders, (2) test scores showing defendant’s third-grade cognition

level, and (3) a corrections officer’s evaluation indicating defendant was reared in

a slum, came early to the attention of juvenile authorities, quit school at 16, and

abused alcohol). 

. And Rompilla

involves the failure of a lawyer to make any inquiry as to a specific, known, and

available piece of evidence, not the failure to make a more open-ended inquiry

about a prior event. See 

125 S. Ct. at 2467 (“Questioning a few more

family members and searching for old records can promise less than looking for a

needle in a haystack, when a lawyer truly has reason to doubt there is any needle

there.”).

       Here, Lane spoke to Miller County officials, an assistant district attorney,

and Johnson himself about the escape and attack. Johnson’s claim, though, is that

Lane unreasonably failed to investigate and present evidence of Johnson’s

motivation for the escape and attack. Johnson first alleged this personal motivation



                                            28
in the state collateral proceedings. His alleged motivation was not evidence in the

possession of the State, nor was it in any public document. It was something

within Johnson’s unique knowledge, and it was something an attorney could

reasonably assume would be shared with him, if important, during an hours-long

discussion of the impact of the escape.12 Consequently, the state habeas court’s

decision that Lane’s investigation of Johnson’s escape and failure to present

evidence of the abuse Johnson received at the Dougherty County jail were not

objectively unreasonable is not contrary to or an unreasonable application of the

facts or clearly established federal law.

B.     Trial Counsel Performance: Evidence of Sheffield’s Death

       Johnson next claims his trial counsel performed deficiently by inadequately



       12
          Johnson argues Lane’s failure to affirmatively ask Johnson why he escaped was
objectively unreasonable performance, citing Wiggins v. Smith, 

, 531-34, 

, 2540-42 (2003), in which the Supreme Court concluded trial counsel’s failure to
investigate the petitioner’s social background beyond a review of the presentence investigation
report and social services records constituted deficient performance, despite the fact that the
petitioner obviously knew his own background and did not volunteer it to counsel. However,
Wiggins expressly relied on the fact that “the evidence counsel uncovered in the social services
records . . . would have led a reasonably competent attorney to investigate further.” 

123 S. Ct. at 2542.
        Here, there was nothing in what Lane had learned about the escape or about Johnson’s
treatment at the Dougherty County jail that would have led any reasonably competent attorney to
ask Johnson why he chose to escape from the Miller County jail. To the contrary, Johnson, as
would any defendant facing capital charges for a brutal killing and with a good chance of being
convicted and sentenced to death, had an obvious reason for escaping: he did not want to be
executed or spend the rest of his life in prison. There was no basis for Lane or any other attorney
in his position to suspect there might be some other reason. The state habeas court’s decision is
not contrary to Wiggins.

                                                29
rebutting the State’s suggestion that Sheffield’s head injury caused Sheffield’s

death eight months later. The state habeas court denied this claim.

      The State sent Lane copies of Sheffield’s medical records about his October

1994 head injury and Sheffield’s June 1995 death certificate, which did not

mention the head injury. Rather, the death certificate listed the cause of death as

congestive heart failure and a stroke. Lane was in Miller County frequently after

the escape, and he saw Sheffield at work supervising community service workers.

Lane reviewed the medical records on Sheffield’s head injury and used them to

cross-examine the State’s witnesses in the penalty phase.

      In concluding there was no deficient performance, the state habeas court

pointed out that Johnson’s trial counsel showed: (1) Sheffield drove himself to the

hospital; (2) Sheffield continued to work after the injury; and (3) Dr. Rummel

admitted she could not say whether there was a causal connection between

Sheffield’s head injury and his death. Lane forced Dr. Rummel to admit that she

ordered a CT scan for Sheffield, which “showed no bleeding” in Sheffield’s brain.

Dr. Rummel also admitted she was “[o]nly indirectly” aware of the CT scan results

– i.e., she knew it showed no bleeding but presumably had not viewed the scan

herself. And Lane elicited testimony from Deputy Houston that after the head

injury Sheffield returned to work for the county and the city and may have directed



                                          30
the public works efforts around the courthouse square.

        Johnson contends Lane’s performance was deficient because he did not (1)

obtain the April 1995 medical records relating to Sheffield’s stroke, and (2) retain

a medical expert to testify there was no causal connection between the head injury

and the stroke.

       We conclude the state habeas court’s decision that Lane’s performance was

not deficient did not involve an unreasonable determination of the facts or

unreasonable application of clearly established federal law. First, there was no

prior notice that Dr. Rummel would testify as she did. Lane said he was surprised.

Second, Lane did take meaningful steps to rebut Dr. Rummel’s suggestion that

Johnson’s attack on Sheffield contributed to his stroke, by eliciting testimony that

Sheffield drove himself to the hospital and returned to work after the injury and

that the CT scan taken after the injury revealed no bleeding. This was effective

rebuttal, particularly given that the majority of Dr. Rummel’s testimony concerned

her opinion that Johnson’s attack on Sheffield was brutal and his injury serious, not

her opinion (which she, and the district attorney, admitted was not certain) that the

injury may have caused Sheffield’s stroke the next year.13 Lane could not deny



       13
          Dr. Rummel testified on direct examination (not cross-examination) that she could not
be entirely sure that Johnson had caused Sheffield’s death. So, from the start, the State’s
causation theory linking Sheffield’s death with Johnson’s attack was weak.

                                               31
that Johnson caused Sheffield’s head injury, and evidence of the attack and escape

was highly aggravating. A reasonable attorney in Lane’s position could have

concluded that to focus too much time and attention on the long-term effect of

Sheffield’s injury would be counterproductive.14

       Third, Sheffield’s death certificate would not put a reasonably competent

attorney on notice of a need to do more discovery or to obtain testimony from a

medical expert. The medical records and CT scan on the head injury, which Lane

had, showed a five- to six-inch laceration on Sheffield’s head but no skull fracture

or intracranial bleeding. The death certificate, which was signed by a doctor other

than Dr. Rummel, indicated Sheffield died of “congestive heart failure” and a

“hemorrhagic stroke – Right brain.” There was no mention of the head injury, plus

Sheffield was 77 years old (which the death certificate indicated). Moreover, the

death certificate showed Sheffield died eight months after the head injury, which

was a significant passage of time from the injury. In addition, Lane knew Sheffield

had returned to work for the county and city during that time period. The mere

inclusion of this document in the State’s production was not enough to require any

       14
         Johnson argues trial counsel should have had Dr. Brian Frist, a pathologist who testified
for the defense in the guilt phase, opine that Sheffield’s death was unrelated to the injury
Johnson inflicted. However, Dr. Frist’s testimony would not have come without cost. Dr. Frist
would have brought Johnson’s attack on Sheffield into the forefront and at a risk that the jury
would view the “he wasn’t hurt that badly” defense as callous. In any event, Johnson’s trial
counsel effectively cross-examined the State’s only witness on this point – Dr. Rummel – and
attacked the State’s suggestion in their closing arguments.

                                                32
reasonably competent attorney in Lane’s position to divert time and resources from

his effort to create lingering doubt to obtain expert testimony on the cause of

Sheffield’s 1995 stroke.

      Under the circumstances, we conclude that the state habeas court did not

unreasonably determine the facts or unreasonably apply clearly established federal

law when it decided that Lane’s penalty phase preparation and performance as to

Sheffield’s death was not objectively unreasonable under prevailing professional

norms.

C.    Trial Counsel Performance: Life History Mitigation Evidence

      Johnson claims his trial counsel were ineffective for failing to adequately

investigate and present mitigating evidence from Johnson’s life history. In the

state habeas proceedings, Johnson proffered testimony from friends and family that

he was capable of love, care, and support; showed particular respect and deference

to the older adults in his family; and supported those with whom he lived. Johnson

also would have had his attorneys focus on his abandonment by his father at age

15, even though he rejoined his father in California two years later. The state

habeas court denied Johnson’s claim, finding trial counsel’s performance was

reasonable:

      Petitioner also alleges that counsel were deficient and that he was
      prejudiced by counsel not presenting more mitigation evidence at trial.

                                          33
       Prior to trial, counsel talked extensively with Petitioner, spoke
with Petitioner’s family members on numerous occasions, and
investigated Petitioner’s childhood. Counsel obtained Petitioner’s
social history, criminal history and medical history. At the sentencing
phase of trial, counsel introduced the testimony of Petitioner’s mother,
a friend of Petitioner’s and two pastors to testify to Petitioner’s good
qualities and characteristics.
       In addition to the heinous facts of the crime and the separate
attack on Deputy Sheffield, the State introduced evidence that
Petitioner had previously been convicted of theft by receiving and
financial transaction card fraud for which he was put on probation;
that, while on probation, Petitioner had to be told a number of times to
quit acting out verbally when he reported to his probation officers; his
probation officers testified that Petitioner had a horrible attitude about
probation and that Petitioner was in the top ten of attitude problems
among the thousands of probationers; Petitioner reported daily drug
use while on probation and failed drug tests; Petitioner refused to keep
probation appointments; Petitioner left the State against probation
rules; Petitioner cursed detention officers while incarcerated; and
Petitioner had verbal altercations with other inmates while
incarcerated.
       Further, in rebuttal to the additional evidence presented by
Petitioner in this habeas proceeding if the same evidence had been
submitted at Petitioner’s trial, the State could have introduced
evidence that Petitioner had a history of physically abusing women,
including Melissa Windows and his ex-wife, Cindy Smith; that Ms.
Smith stated Petitioner had a temper; that Petitioner had been in
previous fights in bars; that Petitioner was noted by the Central State
Hospital Staff as angry, loud and abusive; and that Petitioner was
noted by the jail personnel to be sneaky and quick to anger when he
did not get his way.
       The court finds that the adequacy of a trial attorney’s
investigation into a defendant’s background is judged by a standard of
reasonableness. Counsel were not deficient because trial counsels’
mitigation strategy was reasonable and supported by adequate
investigation and because Petitioner cannot show actual prejudice.
The failure of trial counsel to uncover every possible favorable
witness does not render their performance deficient. The Court

                                   34
       recognizes that post-conviction counsel will almost always be able to
       identify a potential mitigation witness that trial counsel did not
       interview or a record that trial counsel did not obtain.

       After reviewing the evidence, we conclude that the state habeas court’s

conclusion that trial counsel’s performance in investigating and presenting

mitigating evidence from Johnson’s background was not deficient is not based on

an unreasonable determination of the facts or unreasonable application of clearly

established federal law.

       First, the record is clear that trial counsel did investigate Johnson’s

background for mitigating evidence. Although during the first years of the case

Lane focused his efforts on guilt-phase issues and pretrial motions, Lane began his

mitigation investigation in early 1997, more than a year before trial. Lane talked

with Johnson and several of his family members about Johnson’s background.15

       On February 15, 1997, Lane met with Johnson for several hours “to get

some history about [Johnson].” The two discussed Johnson’s school history.

Johnson told Lane about his criminal history, and Lane found “nothing in it that

was very serious.” They also discussed Johnson’s marital, social, employment,

and medical history. Johnson told Lane, among other things, that he: (1) was held



       15
          Additionally, second-chair counsel Jones met with Johnson for nearly five hours in
December 1997 to discuss “mitigation information and family background information.” Jones
testified that “Lane was more in charge of [the mitigation] aspect of the case than I was.”

                                             35
back in third grade because of a reading problem and hyperactivity; (2) did not

participate in sports or activities in Junior High; (3) quit school in tenth grade; (4)

moved about six times from age 16 until the time of the murder; (5) received five

years’ probation for a theft by receiving conviction; (6) had two probation

violations for failing a drug test and leaving the state; (7) was married to his ex-

wife Cindy Smith for two years and lived with her for two years before that; (8)

had used drugs, including marijuana, powder cocaine, methamphetamine, uppers,

downers, heroin (once), and angel dust (once); (9) was arrested for possession of

less than an ounce of marijuana but did not go to court on the charge; (10) received

injuries (including head injuries) in bicycle, motorcycle, and car accidents as a

child and adult; (11) was treated at three different hospitals; (12) was “15 when

[his] parents separated”; and (13) worked at a Jiffy Lube and two restaurants.

Three days later, Lane met with Johnson’s mother and brother and learned of a

time Johnson had defended a woman when a man tried to hurt her.

       Lane got along well with Johnson’s family and he spoke with them about

possible mitigation issues. Lane testified specifically that he discussed Johnson’s

childhood with Johnson’s mother and father.16 Lane also talked a lot with Connie

Givens, who was a friend and former girlfriend of Johnson’s. For example, on


       16
          Lane called Johnson’s mother to testify at the penalty phase, but his father called Lane
before trial and told him he would be unable to attend.

                                                36
November 22, 1994, Lane spent several hours meeting with Givens to discuss

Johnson’s “mental history.” In sum, Johnson’s attorneys clearly investigated

Johnson’s background for mitigation evidence.17

       Second, Lane decided to employ a residual doubt strategy for the penalty

phase. Lane felt Johnson had a chance to be acquitted because of “the

circumstantial nature of the evidence against Mr. Johnson and the lack of

conclusive physical evidence tying him to a homicide.” Lane testified:

       Compared to other death penalty cases where the evidence is just
       overwhelming that the person did it, this was not such a case. You
       would try that case totally differently. You’d forget about did he do
       it. And you start trying it about, well, why did he do it. This case was
       not like that.

Lane believed that if Johnson were convicted, residual doubt would be the only

way to win a life sentence. He testified, “Lingering doubt was the whole thing in

this case. Without it, I didn’t see any hope.”


       17
          In his affidavit, Lane stated that he “neglected to prepare adequately for the sentencing
phase when [he] had the time, and in the end [he] had too little time to develop a coherent and
convincing presentation at sentencing.” Of course, the test for deficient performance is an
objective one, so Lane’s subjective belief that his performance was deficient is far from
determinative. See Windom v. Sec’y, Dep’t of Corr., 

, 1246 n.11 (11th Cir. 2009)
(“Because the adequacy of an attorney’s performance is measured against an objective standard
of reasonableness, the fact that trial counsel admits that his performance was lacking is of little,
if any, consequence.”), cert. denied, 

(2010); Jennings v. McDonough, 

, 1247 (11th Cir. 2007) (“The Strickland standard of objective reasonableness does not
depend on the subjective intentions of the attorney, judgments made in hindsight, or an
attorney’s admission of deficient performance.”). And the objective evidence shows that Lane
engaged in a constitutionally adequate investigation of Johnson’s background, though perhaps
not one that was as comprehensive as in hindsight he now would prefer.

                                                 37
      Lane’s belief in the paramount importance of lingering doubt, and his

consequent decision to focus on guilt-phase (and doubt-creating) issues for most of

the pretrial period, was not objectively unreasonable. As the state habeas court

noted, “the state had no eyewitness to the crime itself and was forced to present a

circumstantial evidence case” and Lane’s chance to provide sufficient mitigating

evidence at the penalty phase “was made incredibly more difficult” by Johnson’s

escape from the Miller County jail and attack on Sheffield.

      Johnson argues that the scope of trial counsel’s investigation into mitigating

evidence from Johnson’s life history was objectively unreasonable. However, as

the Supreme Court stated in Van Hook while rejecting a similar argument:

      This is not a case in which the defendant’s attorneys failed to act
      while potentially powerful mitigating evidence stared them in the
      face, cf. 

, 

, or would have
      been apparent from documents any reasonable attorney would have
      obtained, cf. Rompilla v. Beard, 

, 389-393, 

, 

(2005). It is instead a case, like Strickland
      itself, in which defense counsel’s “decision not to seek more”
      mitigating evidence from the defendant’s background “than was
      already in hand” fell “well within the range of professionally
      reasonable judgments.”

Van 

. Here, trial counsel investigated Johnson’s

background, in particular his social, education, criminal, and medical history, but

found no “potentially powerful mitigating evidence.” Under the circumstances,

reasonable attorneys could have chosen, as trial counsel did here, to focus most of

                                          38
their time and effort on investigations supporting residual doubt.

      As to presentation, trial counsel called four witnesses in the penalty phase: a

minister and teacher who each spoke about Johnson’s good character; a friend of

Johnson’s who testified about Johnson’s love of cabinetmaking and dance; and

Johnson’s mother, who pled for mercy on her son’s behalf. This testimony served

to humanize Johnson to the jury.

      Moreover, had Johnson’s trial counsel emphasized Johnson’s good

character, as Johnson now contends they should have, the State would have had

greater incentive to introduce rebuttal evidence (which it did not introduce at trial)

that not only would have undercut the good-character evidence, but also would

have dealt a powerful blow to the defense’s core penalty-phase strategy of creating

lingering doubt. In particular, the State could have introduced evidence that

Johnson had a temper and physically abused his ex-wife Cynthia Smith, his ex-

girlfriend Melissa Windows, and, on one occasion, one of Smith’s daughters.

Given this potential rebuttal evidence, and the aggravating evidence of (1) the

circumstances of Sizemore’s murder, rape, and mutilation, and (2) Johnson’s

escape and attack on Sheffield, it was reasonable for counsel to present evidence of

Johnson’s childhood, hobbies, and a mercy plea from Johnson’s mother in lieu of a

full-bore good-character strategy. The new mitigation evidence would have risked



                                          39
the State presenting specific rebuttal evidence that would have damaged Johnson’s

character. See Wood v. Allen, 

, 1313 (11th Cir. 2008) (“[W]e have

rejected prejudice arguments where mitigation evidence . . . would have opened the

door to damaging evidence.”), aff’d, — U.S. —, 

(2010).18

D.     Trial Counsel Performance: Expert Future-Dangerousness Testimony

       Johnson also claims trial counsel were deficient in the penalty phase by not

presenting expert testimony and statistics on the issue of Johnson’s future

dangerousness. In the state habeas proceedings, Johnson introduced expert and

statistical evidence about prison trends and future dangerousness of lifers in prison.

For example, forensic psychologist Dr. Thomas Reidy testified about studies

showing that (1) the offense of conviction, prior convictions, and escape history

are only weakly correlated with prison violence, (2) prisoners are less likely to

misbehave as they age, and (3) prisoners with long-term sentences are more likely

to successfully adjust to prison. Dr. Reidy opined that Georgia would classify


       18
          Johnson argues that his trial counsel already introduced evidence of his good character
at the penalty phase (just not enough), and the State did not introduce this rebuttal evidence, so it
is not appropriate to assume the State would do so in the face of the proffered evidence. This
ignores the fact that when, in the state habeas case, Johnson proffered such evidence, the State
introduced the rebuttal evidence of, inter alia, Johnson’s domestic abuse and violent temper.
Thus, it is reasonable to conclude the odds were good that had Johnson’s counsel introduced the
proffered mitigation evidence at trial, the State would have introduced its proffered rebuttal
evidence. The point is that if Johnson’s counsel had focused their penalty-phase presentation on
life history mitigation evidence (as opposed to lingering doubt) and called a lot more witnesses
as to Johnson’s good character as an adult, the State would have done more on this score too – as
happened in the state habeas hearing.

                                                 40
Johnson as a “maximum security” or “hi-max” prisoner and has safeguards to

minimize the risk posed by such potentially dangerous prisoners, and thus Johnson

was less likely to be violent in prison.

      Former corrections commissioner and warden James Aiken testified that

Georgia prisoners are effectively inhibited from committing violent infractions by

the structure and security measures of the prison environment. Aiken testified that

Johnson would receive a “maximum” or “hi-max” security classification, “meaning

that he would be subjected to the most stringent security measures available,

including highly restrictive confinement and close supervision.” If Johnson were

“deemed a problem inmate inclined toward misconduct, adequate measures would

be taken to segregate him from the general prison population and restrict his

movement, thereby negating the opportunity and ability to commit misconduct.”

Aiken and Dr. Reidy listed some of the high-level security and prophylactic

measures employed in state prisons: housing in high-security facilities equipped

with electronic perimeter detection systems, closed circuit television cameras, and

listening devices; single cell placement; and administrative segregation.

      Sociologist Dr. Michael Radelet testified that most prisoners convicted of

capital murder but sentenced to long-term imprisonment adjust satisfactorily to

prison. Dr. Radelet agreed with Dr. Reidy’s and Aiken’s conclusions and opined



                                           41
that Johnson’s probability for a satisfactory adjustment to prison would be greater

than most because he was older than most offenders and had no prior convictions

for violent crimes.

       Dr. Tim Carr, a Georgia corrections department statistician, compiled

statistics regarding infraction rates between long-term inmates and the general

inmate population. Dr. Carr testified that long-term inmates generally have lower

rates of infractions because they are housed in more secure conditions. But Dr.

Carr testified that the statistics showed only minor variations and were not

conclusive, as follows:

       Q      Putting what you said about very little difference between the
       numbers aside, or in your opinion, could I go through and basically
       pick and choose things to support whatever position I wanted?
       A      Yes. If you wanted to say, for instance, that the general
       population inmates were better than, for example, lifers, you could
       pick out many examples from the reams of statistics that I produced,
       or you could make exactly the opposite case and pull out numerous
       examples in the opposite direction.
       ....
       [I]f you were to go carefully through those printouts and put a poker
       chip over here in this pile every time you found where lifers were
       better than average inmates, put it in this pile, when you find the
       opposite, you put it in this pile, by the time you had finished you
       would have, roughly, equal piles of poker chips.
       Q       Okay. So again, I can make these numbers say just about
       anything I want them to?
       A That’s correct.[19]


       19
         We reject Johnson’s claim that the state habeas court’s conclusion that the statistics
were easily manipulable was an unreasonable determination of the facts.

                                                42
      As to the expert future dangerousness testimony, the state habeas court noted

that: (1) “Petitioner’s habeas experts testified that the prison system would find

that Petitioner was a high risk and dangerous and therefore place him in maximum

security which would ‘control’ his dangerousness”; (2) “[e]xperts conceded that

Petitioner would be classified as ‘maximum’ or ‘hi-max’ security level because of

his crime and past escape”; and (3) “[t]he statistician for the Georgia Department

of Corrections who compiled the statistics used by Petitioner’s experts in this case

testified that the data, over all, showed minor variances, was easily manipulated

and could be used to support almost any conclusion.”

      Before Johnson’s trial, Lane actually had used Dr. Radelet as a witness on

future dangerousness in another capital case. Thus, Lane knew about the

availability of experts who would testify that a convicted murderer serving a long

sentence will be classified as maximum security and thus will be in a highly

restricted confinement under close supervision, will age and adjust to prison life,

and will not present a danger in prison. Even though Lane contacted Dr. Radelet

about Johnson’s case and got funds to hire him, Lane never called Dr. Radelet as a

witness. The state habeas court found that Lane made a strategic decision not to

call Dr. Radelet in Johnson’s case, but that fact finding is not supported by the

record. Although Lane obtained the funds to pay for Dr. Radelet’s 15 hours’



                                          43
preparation time approximately three weeks before he was to testify, Lane testified

Dr. Radelet had a busy teaching schedule and was too busy to get ready in that

time period. Thus, we examine Johnson’s future-dangerousness claim de novo on

the basis of whether a reasonably competent trial counsel is deficient for not

presenting this type of future dangerousness testimony that counsel knew experts

would give.

      The State makes several strong arguments about how a trial counsel’s not

presenting this expert future dangerousness testimony “falls within the wide range

of reasonable professional assistance.” 

, 104 S. Ct. at

2065. First, the experts’ opinions that Johnson would be unlikely to harm guards

or fellow inmates in prison were based in large part on the fact that prison

authorities would classify him as a maximum security inmate based on his murder

conviction and escape history. Their opinions did not focus on a mitigating lack of

propensity to violence in Johnson’s character, but rather on the fact that the

structured, maximum security prison setting likely would control Johnson and not

permit him to act on any desires to escape or commit acts of violence. They also

serve to emphasize the point, not favorable to Johnson, that the prison system

would classify him as a high security inmate because it would consider him to

present a high risk of misconduct or escape.



                                          44
       Second, the statistics upon which Johnson’s experts rely could be read either

way. In other words, as statistician Dr. Carr testified, the statistics showed only

minor variances between the prison misconduct rates of life-sentenced prisoners

and the general population, were not conclusive, and could be used to support

either position. Additionally, Carr’s statistics were of a general nature that would

have had little, if any, mitigating effect.

       Third, a reasonable attorney could simply conclude that the experts’ core

opinion – that statistical data about other inmates proved Johnson presented a low

risk of future harm in spite of his having recently attacked a jailer and escaped –

was weak and would not persuade Johnson’s jurors.

       For the reasons argued by the State, we conclude that a reasonable attorney

could decide that under the particular facts of this case the future dangerousness

testimony would not be helpful.20 Thus, Johnson has not shown deficient


       20
          We note that this is not a case in which the State introduced expert testimony about the
defendant’s future dangerousness and the defendant was denied an opportunity to directly rebut
that testimony with his own expert. See Clisby v. Jones, 

, 929 n.7 (11th Cir. 1992).
Nor is this a case like Williams v. Taylor, 

, 

(2000), cited by
Johnson, in which the State’s two experts testified there was a “high probability” that Williams
posed a serious continuing security threat and defendant’s attorneys failed to elicit: (1) evidence
of Williams’s “nightmarish childhood,” including criminal neglect by his parents, severe abuse
by his father, and his time in an abusive foster home; (2) that correctional officers were willing
to testify Williams would not pose a danger while incarcerated; (3) that Williams received prison
commendations for breaking up a drug ring and returning an officer’s wallet; and (4) an
admission from the State’s testifying future dangerousness experts that the defendant would not
pose a threat in the future if kept in a structured environment. 

373 & n.4, 

& n.4, 1514-15.

                                                45
performance as to this claim.

E.     Prejudice

       The state habeas court found that Johnson did not satisfy the prejudice prong

as to any ineffective trial counsel claims. The state habeas court found the expert

testimony about future dangerousness would have been unpersuasive to any juror

given Johnson had stabbed Sizemore 41 times, had sexually assaulted her with a

tree limb, and had escaped from jail and assaulted a 76-year-old guard with a gun,

stating:

       In light of the evidence that Petitioner had stabbed Angela Sizemore
       41 times, vaginally assaulted her with a tree limb while she was still
       alive, subsequently escaped from the Miller County Jail by taking a 76
       year-old guard’s gun and assaulting the guard with the gun causing an
       injury requiring 21 staples, this additional evidence to attempt to show
       Petitioner would not be dangerous in the future in a prison setting
       through expert opinion testimony would have been unpersuasive to
       the jury and would not with reasonable probability have changed the
       outcome of Petitioner’s case in the face of such irrefutable evidence of
       the defendant’s apparent ability to justify himself in an aggravated
       assault on an elderly detention officer in order to illegally gain his
       own freedom for whatever reason.

The state habeas court also stressed that Johnson’s trial attorneys had no evidence

to refute Johnson’s actual commission of the aggravated assault on the jail guard,

finding:

       The Petitioner submits no argument that there was any evidence that
       his trial attorneys had available to refute the actual commission of the
       aggravated assault and escape. The court does not find that the

                                         46
      evidence as to the extent of the injury to the jailer and the testimony of
      the pathologists contradicting the treating physician would . . . have
      deterred a juror who had already found the defendant guilty of raping,
      torturing and murdering Ms. Sizemore from imposing the death
      penalty on petitioner, and the court does not find counsel ineffective
      in not offering such unpersuasive evidence.

The state habeas court noted that if Johnson’s additional evidence was presented,

the State would have presented additional aggravating evidence, such as how

Johnson had physically attacked his ex-girlfriend and ex-wife, had been in fights in

bars, and was described by jail officials as sneaky and quick to anger when he did

not get his way.

      Considering “collectively” all of the evidence at the state habeas hearing, the

state habeas court found no prejudice:

      The court further finds that even if the additional evidence submitted
      by Petitioner in mitigation had been presented at trial, Petitioner must
      establish prejudice. This court is unable to conclude under the
      relevant burden of proof that the evidence presented at the state
      habeas corpus hearing, considered collectively, in light of the
      additional, potential aggravating evidence available to the State would
      with reasonable probability have changed the outcome of Petitioner’s
      trial.

(Emphasis omitted.)

      We conclude that the state habeas court’s decision as to the prejudice prong

of Johnson’s ineffective trial counsel claims is not contrary to, nor based on an




                                          47
unreasonable application of, clearly established federal law.21

       Most if not all of the mitigation evidence Johnson proffers is weak when

compared to the evidence in aggravation. The jury found Johnson guilty of the

brutal, torturous murder of Angela Sizemore. The nature of the crime is powerful

evidence in aggravation – Johnson stabbed Sizemore 41 times and sexually

assaulted her with a pecan tree limb, all while she was still alive, and then dragged

and dumped her body. The victim impact evidence is powerful too, because

Johnson’s killing of Sizemore left a four-year-old girl motherless and devastated

Sizemore’s entire family. Sizemore’s sister testified her own death would be easier

than living with daily thoughts of what Sizemore’s last moments were like.

Further, the evidence was undisputed that while awaiting trial for Sizemore’s

murder, Johnson attacked a 76-year-old jailer to escape from jail, an attack that

required 21 staples to close the jailer’s head wound. The state habeas court

reasonably concluded that Johnson’s own conduct made his trial counsel’s task at

the penalty phase “incredibly more difficult.”

       Moreover, much of Johnson’s mitigating evidence would be contradicted or

undercut by other evidence. For example, hospital treatment records show that




       21
         We also reject Johnson’s arguments that the state habeas court applied the wrong legal
standard or based its decision on an unreasonable determination of the facts.

                                               48
Johnson denied being the victim of a sexual assault at the Dougherty County jail.22

While Johnson claimed he was physically abused, the only documented instance is

a June 1994 attack for which Johnson was taken to the emergency room and

diagnosed with swelling in his left wrist, a knot on his forehead, and bruises to his

neck, arms, and face. Although Johnson claimed he escaped for fear of returning

to the Dougherty County jail, Johnson was in the Miller County jail at the time of

his escape and there were no plans at the time to transfer him back to Dougherty

County. Indeed, even after his escape, Johnson was returned to the Miller County

jail and he never told attorney Lane of this alleged fear.23 More fundamentally,

even if the jury accepted Johnson’s alleged reason for his escape and that his attack

on Sheffield was unrelated to the stroke, the undisputed evidence still reveals

Johnson as a man willing to savagely beat a 76-year-old jailer to save himself.

       Johnson’s other categories of proposed mitigation suffer from similar flaws.

Although Johnson’s future dangerousness experts opined that they believed


       22
          In July 1995, Johnson underwent a psychological assessment at Central State Hospital,
as part of which Johnson completed a “comprehensive assessment” questionnaire. One of the
questions was, “Have you ever been sexually or physically abused? Explain.” Johnson wrote
“No” next to “sexually,” and next to “physically” he wrote, “physically abused in Dougherty Co.
jail.”
       23
         The jury easily could have rejected the notion that Johnson had a reasonable belief that
he was going to be transferred, especially since after the escape he went back to the Miller
County jail and ultimately went to the new Dougherty County jail facility when it opened in
January 1995. Alternatively, the jury could have concluded that if Johnson feared a transfer, he
simply could have asked Lane to make sure he stayed in the Miller County jail.

                                                49
Johnson would likely adjust well to a prison setting and would not likely harm

other people in prison, their testimony was based in part on the fact that Johnson

would be classified by the prison system as a maximum security prisoner because

of his murder offense and escape history.24 In other words, Johnson would be

unlikely to harm others not necessarily because he is unwilling to do so but

because he would be watched and confined enough that he would be unlikely to be

able to. Such testimony is obviously less mitigating than testimony that a

defendant would not be dangerous because of his character or other intrinsic

factors. Moreover, as the state habeas court noted, Johnson’s experts’ opinions

were based on an analysis of prison statistics that, according to the statistician who

compiled them, showed only very small differences and could be used to support

almost any position.

       Similarly, Johnson’s proffered life history mitigation is utterly devoid of the

power or emotional heft of the evidence present in cases where the Supreme Court

has determined that the prejudice prong was satisfied. Johnson was not sexually or

physically abused or deprived of food by his parents, and they did not separate


       24
         Although Johnson’s experts believed the prison system’s classification would
“overcompensate” for Johnson’s actual risk profile, that belief rested on, inter alia, the experts’
discounting of Johnson’s escape history as a “unique” event permitted by Sheffield’s negligent
behavior and motivated by Johnson’s “reason[able]” fear of being transferred back to the
Dougherty County jail. Jurors may have chosen not to discount, as Johnson’s experts did,
Johnson’s actual history of violent escape.

                                                 50
until Johnson was fifteen. Compare Porter v. McCollum, — U.S. —, 

, 454 (2009) (finding evidence of Porter’s heroic military service and resulting

emotional scars, brain abnormality, and childhood physical abuse was the “kind of

troubled history we have declared relevant to assessing a defendant’s moral

culpability”); 

, 125 S. Ct. at 2468-69 (finding

prejudice prong satisfied by evidence that Rompilla’s parents were alcoholics, his

father frequently beat his mother and bragged about his infidelity, his father beat

Rompilla and locked him and his brother in an excrement-filled dog pen, Rompilla

slept in an unheated attic, and he was given no clothes and went to school in rags);

, 123 S. Ct. at 2542 (noting evidence, inter alia, of

“severe privation and abuse in the first six years of . . . life,” and “physical torment,

sexual molestation, and repeated rape during . . . subsequent years in foster care”);

, 120 S. Ct. at 1514-15 (finding counsel prejudiced

Williams by not presenting graphic evidence of defendant’s “nightmarish

childhood” that included his parents’ imprisonment for criminal neglect of him and

his siblings, his severe and frequent beatings by his father, and his commitment to

an abusive foster home). Nor would it enable the jury to “understand [Johnson]

and his horrendous acts.” Sears v. Upton, 561 U.S. —, 

, 3264

(2010).



                                           51
      Instead, Johnson’s proffered life history mitigation of Johnson’s good

character as an adult stood in stark contrast to the brutal nature of Sizemore’s

murder and Johnson’s admitted aggravated assault on Deputy Sheffield, and may

well have prompted the State to introduce evidence of Johnson’s violent temper

and physical abuse of his ex-wife Cynthia Smith and his ex-girlfriend Melissa

Windows, which not only would have undercut Johnson’s good-character

mitigation but also would have devastated his residual doubt theory. If the

additional life history evidence was admitted, the State would have had more

incentive to introduce the rebuttal evidence listed by the state habeas court: (1)

“that Petitioner had a history of physically abusing women, including Melissa

Windows and his ex-wife, Cindy Smith”; (2) “that Ms. Smith stated Petitioner had

a temper”; (3) “that Petitioner had been in previous fights in bars”; (4) “that

Petitioner was noted by the Central State Hospital Staff as angry, loud and

abusive”; and (5) “that Petitioner was noted by the jail personnel to be sneaky and

quick to anger when he did not get his way.”

      On prejudice grounds, this case is closer to Wong v. Belmontes, 558 U.S.

—, 

, 387-90 (2009), in which the Supreme Court found no prejudice

from counsel’s failure to present (1) cumulative evidence of Belmontes’s difficult

childhood and good character, (2) expert opinion that he was likely to have a



                                          52
nonviolent adjustment to a prison setting that would have opened the door to

evidence Belmontes had committed another murder, and (3) evidence of

Belmontes’s emotional instability and impaired planning and reasoning ability that

was undercut by the calculated nature of the murder and Belmontes’s subsequent

bragging about it.

      In sum, Johnson’s state habeas evidence and the mitigation testimony at the

penalty phase, considered collectively, are not nearly strong enough when weighed

against the extreme evidence in aggravation for us to conclude that Johnson has

shown prejudice. Contrary to Johnson’s assertions, this is a case in which “the new

evidence would barely have altered the sentencing profile presented” to Johnson’s

jury. 

(quotation marks omitted). It was thus not

unreasonable for the state habeas court to conclude that Johnson had failed to show

a reasonable probability that he would receive a different sentence. Accordingly,

the state habeas court did not unreasonably determine the facts or unreasonably

apply established Supreme Court precedent when it concluded Johnson did not

satisfy the prejudice prong as to his ineffective trial counsel claims.

                                 IX. CONCLUSION

      We affirm the district court’s denial of Johnson’s § 2254 petition.

      AFFIRMED.



                                           53