Cite as: 590 U. S. ____ (2020) 1
Per Curiam
SUPREME COURT OF THE UNITED STATES
TERENCE TRAMAINE ANDRUS v. TEXAS
ON PETITION FOR WRIT OF CERTIORARI TO THE
COURT OF CRIMINAL APPEALS OF TEXAS
No. 18–9674. Decided June 15, 2020
PER CURIAM.
Death-sentenced petitioner Terence Andrus was six
years old when his mother began selling drugs out of the
apartment where Andrus and his four siblings lived. To
fund a spiraling drug addiction, Andrus’ mother also turned
to prostitution. By the time Andrus was 12, his mother reg-
ularly spent entire weekends, at times weeks, away from
her five children to binge on drugs. When she did spend
time around her children, she often was high and brought
with her a revolving door of drug-addicted, sometimes phys-
ically violent, boyfriends. Before he reached adolescence,
Andrus took on the role of caretaker for his four siblings.
When Andrus was 16, he allegedly served as a lookout
while his friends robbed a woman. He was sent to a juvenile
detention facility where, for 18 months, he was steeped in
gang culture, dosed on high quantities of psychotropic
drugs, and frequently relegated to extended stints of soli-
tary confinement. The ordeal left an already traumatized
Andrus all but suicidal. Those suicidal urges resurfaced
later in Andrus’ adult life.
During Andrus’ capital trial, however, nearly none of this
mitigating evidence reached the jury. That is because An-
drus’ defense counsel not only neglected to present it; he
failed even to look for it. Indeed, counsel performed virtu-
ally no investigation of the relevant evidence. Those fail-
ures also fettered the defense’s capacity to contextualize or
counter the State’s evidence of Andrus’ alleged incidences
of past violence.
Only years later, during an 8-day evidentiary hearing in
2 ANDRUS v. TEXAS
Per Curiam
Andrus’ state habeas proceeding, did the grim facts of An-
drus’ life history come to light. And when pressed at the
hearing to provide his reasons for failing to investigate An-
drus’ history, Andrus’ counsel offered none.
The Texas trial court that heard the evidence recom-
mended that Andrus be granted habeas relief and receive a
new sentencing proceeding. The court found the abundant
mitigating evidence so compelling, and so readily available,
that counsel’s failure to investigate it was constitutionally
deficient performance that prejudiced Andrus during the
punishment phase of his trial. The Texas Court of Criminal
Appeals disagreed. It concluded without explanation that
Andrus had failed to satisfy his burden of showing ineffec-
tive assistance under Strickland v. Washington, 466 U. S.
668 (1984).
We conclude that the record makes clear that Andrus has
demonstrated counsel’s deficient performance under Strick-
land, but that the Court of Criminal Appeals may have
failed properly to engage with the follow-on question
whether Andrus has shown that counsel’s deficient perfor-
mance prejudiced him. We thus grant Andrus’ petition for
a writ of certiorari, vacate the judgment of the Texas Court
of Criminal Appeals, and remand the case for further pro-
ceedings not inconsistent with this opinion.
I
A
In 2008, 20-year-old Terence Andrus unsuccessfully at-
tempted a carjacking in a grocery-store parking lot while
under the influence of PCP-laced marijuana. During the
bungled attempt, Andrus fired multiple shots, killing car
owner Avelino Diaz and bystander Kim-Phuong Vu Bui.
The State charged Andrus with capital murder.
At the guilt phase of trial, Andrus’ defense counsel de-
clined to present an opening statement. After the State
rested its case, the defense immediately rested as well. In
Cite as: 590 U. S. ____ (2020) 3
Per Curiam
his closing argument, defense counsel conceded Andrus’
guilt and informed the jury that the trial would “boil down
to the punishment phase,” emphasizing that “that’s where
we are going to be fighting.” 45 Tr. 18. The jury found An-
drus guilty of capital murder.
Trial then turned to the punishment phase. Once again,
Andrus’ counsel presented no opening statement. In its 3-
day case in aggravation, the State put forth evidence that
Andrus had displayed aggressive and hostile behavior
while confined in a juvenile detention center; that Andrus
had tattoos indicating gang affiliations; and that Andrus
had hit, kicked, and thrown excrement at prison officials
while awaiting trial. The State also presented evidence ty-
ing Andrus to an aggravated robbery of a dry-cleaning busi-
ness. Counsel raised no material objections to the State’s
evidence and cross-examined the State’s witnesses only
briefly.
When it came to the defense’s case in mitigation, counsel
first called Andrus’ mother to testify. The direct examina-
tion focused on Andrus’ basic biographical information and
did not reveal any difficult circumstances in Andrus’ child-
hood. Andrus’ mother testified that Andrus had an “excel-
lent” relationship with his siblings and grandparents. 49
id., at 52, 71. She also insisted that Andrus “didn’t have
access to” “drugs or pills in [her] household,” and that she
would have “counsel[ed] him” had she found out that he was
using drugs. Id., at 67, 79.
The second witness was Andrus’ biological father, Mi-
chael Davis, with whom Andrus had lived for about a year
when Andrus was around 15 years old. Davis had been in
and out of prison for much of Andrus’ life and, before he
appeared to testify, had not seen Andrus in more than six
years. The bulk of Davis’ direct examination explored such
topics as Davis’ criminal history and his relationship with
Andrus’ mother. Toward the end of the direct examination,
counsel elicited testimony that Andrus had been “good
4 ANDRUS v. TEXAS
Per Curiam
around [Davis]” during the 1-year period he had lived with
Davis. 50 id., at 8.
Once Davis stepped down, Andrus’ counsel informed the
court that the defense rested its case and did not intend to
call any more witnesses. After the court questioned counsel
about this choice during a sidebar discussion, however,
counsel changed his mind and decided to call additional wit-
nesses.
Following a court recess, Andrus’ counsel called Dr. John
Roache as the defense’s only expert witness. Counsel’s
terse direct examination focused on the general effects
of drug use on developing adolescent brains. On cross-
examination, the State quizzed Dr. Roache about the rele-
vance and purpose of his testimony, probing pointedly
whether Dr. Roache “drove three hours from San Antonio
to tell the jury . . . that people change their behavior when
they use drugs.” 51 id., at 21.
Counsel next called James Martins, a prison counselor
who had worked with Andrus. Martins testified that An-
drus “started having remorse” in the past two months and
was “making progress.” Id., at 35. On cross-examination,
the State emphasized that Andrus’ feelings of remorse had
manifested only recently, around the time trial began.
Finally, Andrus himself testified. Contrary to his
mother’s depiction of his upbringing, he stated that his
mother had started selling drugs when he was around six
years old, and that he and his siblings were often home
alone when they were growing up. He also explained that
he first started using drugs regularly around the time he
was 15. All told, counsel’s questioning about Andrus’ child-
hood comprised four pages of the trial transcript. The State
on cross declared, “I have not heard one mitigating circum-
stance in your life.” Id., at 60.
The jury sentenced Andrus to death.
Cite as: 590 U. S. ____ (2020) 5
Per Curiam
B
After an unsuccessful direct appeal, Andrus filed a state
habeas application, principally alleging that his trial coun-
sel was ineffective for failing to investigate or present avail-
able mitigation evidence. During an 8-day evidentiary
hearing, Andrus presented what the Texas trial court char-
acterized as a “tidal wave of information . . . with regard to
mitigation.” 7 Habeas Tr. 101.
The evidence revealed a childhood marked by extreme ne-
glect and privation, a family environment filled with vio-
lence and abuse. Andrus was born into a neighborhood of
Houston, Texas, known for its frequent shootings, gang
fights, and drug overdoses. Andrus’ mother had Andrus,
her second of five children, when she was 17. The children’s
fathers never stayed as part of the family. One of them
raped Andrus’ younger half sister when she was a child.
The others—some physically abusive toward Andrus’
mother, all addicted to drugs and carrying criminal histo-
ries—constantly flitted in and out of the picture.
Starting when Andrus was young, his mother sold drugs
and engaged in prostitution. She often made her drug sales
at home, in view of Andrus and his siblings. She also ha-
bitually used drugs in front of them, and was high more of-
ten than not. In her frequently disoriented state, she would
leave her children to fend for themselves. Many times,
there was not enough food to eat.
After her boyfriend was killed in a shooting, Andrus’
mother became increasingly dependent on drugs and ne-
glectful of her children. As a close family friend attested,
Andrus’ mother “would occasionally just take a week or a
weekend and binge [on drugs]. She would get a room some-
where and just go at it.” 13 Habeas Tr., Def. Exh. 13, p. 2.
With the children often left on their own, Andrus as-
sumed responsibility as the head of the household for his
four siblings, including his older brother with special needs.
Andrus was around 12 years old at the time. He cleaned
6 ANDRUS v. TEXAS
Per Curiam
for his siblings, put them to bed, cooked breakfast for them,
made sure they got ready for school, helped them with their
homework, and made them dinner. According to his sib-
lings, Andrus was “a protective older brother” who “kept on
to [them] to stay out of trouble.” Id., Def. Exh. 18, p. 1. An-
drus, by their account, was “very caring and very loving,”
“liked to make people laugh,” and “never liked to see people
cry.” Ibid.; id., Def. Exh. 9, p. 1. While attempting to care
for his siblings, Andrus struggled with mental-health is-
sues: When he was only 10 or 11, he was diagnosed with
affective psychosis.
At age 16, Andrus was sentenced to a juvenile detention
center run by the Texas Youth Commission (TYC), for al-
legedly “serv[ing] as the ‘lookout’ ” while he and his friends
robbed a woman of her purse. 10 Habeas Tr., State Exh.
16, p. 9; 13 id., Def. Exh. 4, p. 4 (“[R]ecords indicate[d that]
Andrus served as the lookout”); 3 id., at 273–274; 5 id., at
206.1 While in TYC custody, Andrus was prescribed high
doses of psychotropic drugs carrying serious adverse side
effects. He also spent extended periods in isolation, often
for purported infractions like reporting that he had heard
voices telling him to do bad things. TYC records on Andrus
noted multiple instances of self-harm and threats of suicide.
After 18 months in TYC custody, Andrus was transferred
to an adult prison facility.
Not long after Andrus’ release from prison at age 18, An-
drus attempted the fatal carjacking that resulted in his cap-
ital convictions. While incarcerated awaiting trial, Andrus
tried to commit suicide. He slashed his wrist with a razor
——————
1 The dissent states that the victim identified Andrus as the individual
holding the gun, post, at 5 (opinion of ALITO, J.), but in fact, the victim
testified at Andrus’ trial that she did not and could not identify faces or
individuals, see 4 Tr. 17, 19–20. The dissent also claims that “the victim
matched Andrus’s clothing to the gunman’s,” post, at 5, n. 1, but neglects
to mention that the victim described at least two individuals as wearing
such clothing, see 46 Tr. 25–27.
Cite as: 590 U. S. ____ (2020) 7
Per Curiam
blade and used his blood to smear messages on the walls,
beseeching the world to “[j]ust let [him] die.” 31 id., Def.
Exh. 122–A, ANDRUS–SH 4522.
After considering all the evidence at the hearing, the
Texas trial court concluded that Andrus’ counsel had been
ineffective for “failing to investigate and present mitigating
evidence regarding [Andrus’] abusive and neglectful child-
hood.” App. to Pet. for Cert. 36. The court observed that
the reason Andrus’ jury did not hear “relevant, available,
and persuasive mitigating evidence” was that trial counsel
had “fail[ed] to investigate and present all other mitigating
evidence.” Id., at 36–37. The court explained that “there
[is] ample mitigating evidence which could have, and
should have, been presented at the punishment phase of
[Andrus’] trial.” Id., at 36. For that reason, the court con-
cluded that counsel had been constitutionally ineffective,
and that habeas relief, in the form of a new punishment
trial, was warranted. Id., at 37, 42.
C
The Texas Court of Criminal Appeals rejected the trial
court’s recommendation to grant habeas relief. In an un-
published per curiam order, the Court of Criminal Appeals
concluded without elaboration that Andrus had “fail[ed] to
meet his burden under Strickland v. Washington, 466 U. S.
668 (1984), to show by a preponderance of the evidence that
his counsel’s representation fell below an objective stand-
ard of reasonableness and that there was a reasonable prob-
ability that the result of the proceedings would have been
different but for counsel’s deficient performance.” App. to
Pet. for Cert. 7–8. A concurring opinion reasoned that, even
if counsel had provided deficient performance under Strick-
land, Andrus could not show that counsel’s deficient perfor-
mance prejudiced him.
Andrus petitioned for a writ of certiorari. We grant the
8 ANDRUS v. TEXAS
Per Curiam
petition, vacate the judgment of the Texas Court of Crimi-
nal Appeals, and remand for further proceedings not incon-
sistent with this opinion. The evidence makes clear that
Andrus’ counsel provided constitutionally deficient perfor-
mance under Strickland. But we remand so that the Court
of Criminal Appeals may address the prejudice prong of
Strickland in the first instance.
II
To prevail on a Sixth Amendment claim alleging ineffec-
tive assistance of counsel, a defendant must show that his
counsel’s performance was deficient and that his counsel’s
deficient performance prejudiced him. Strickland, 466
U. S., at 688, 694. To show deficiency, a defendant must
show that “counsel’s representation fell below an objective
standard of reasonableness.” Id., at 688. And to establish
prejudice, a defendant must show “that there is a reasona-
ble probability that, but for counsel’s unprofessional errors,
the result of the proceeding would have been different.” Id.,
at 694.
A
“It is unquestioned that under prevailing professional
norms at the time of [Andrus’] trial, counsel had an ‘obliga-
tion to conduct a thorough investigation of the defendant’s
background.’ ” Porter v. McCollum, 558 U. S. 30, 39 (2009)
(per curiam) (quoting Williams v. Taylor, 529 U. S. 362, 396
(2000)). Counsel in a death-penalty case has “ ‘a duty to
make reasonable investigations or to make a reasonable de-
cision that makes particular investigations unnecessary.’ ”
Wiggins v. Smith, 539 U. S. 510, 521 (2003) (quoting Strick-
land, 466 U. S., at 691). “ ‘In any ineffectiveness case, a par-
ticular decision not to investigate must be directly assessed
for reasonableness in all the circumstances, applying a
heavy measure of deference to counsel’s judgments.’ ” Wig-
gins, 539 U. S., at 521–522.
Cite as: 590 U. S. ____ (2020) 9
Per Curiam
Here, the habeas record reveals that Andrus’ counsel fell
short of his obligation in multiple ways: First, counsel per-
formed almost no mitigation investigation, overlooking vast
tranches of mitigating evidence. Second, due to counsel’s
failure to investigate compelling mitigating evidence, what
little evidence counsel did present backfired by bolstering
the State’s aggravation case. Third, counsel failed ade-
quately to investigate the State’s aggravating evidence,
thereby forgoing critical opportunities to rebut the case in
aggravation. Taken together, those deficiencies effected an
unconstitutional abnegation of prevailing professional
norms.
1
To assess whether counsel exercised objectively reasona-
ble judgment under prevailing professional standards, we
first ask “whether the investigation supporting counsel’s
decision not to introduce mitigating evidence of [Andrus’]
background was itself reasonable.” Id., at 523 (emphasis
deleted); see also id., at 528 (considering whether “the scope
of counsel’s investigation into petitioner’s background” was
reasonable); Porter, 558 U. S., at 39. Here, plainly not. Alt-
hough counsel nominally put on a case in mitigation in that
counsel in fact called witnesses to the stand after the pros-
ecution rested, the record leaves no doubt that counsel’s in-
vestigation to support that case was an empty exercise.
To start, counsel was, by his own admissions at the ha-
beas hearing, barely acquainted with the witnesses who
testified during the case in mitigation. Counsel acknowl-
edged that the first time he met Andrus’ mother was when
she was subpoenaed to testify, and the first time he met
Andrus’ biological father was when he showed up at the
courthouse to take the stand. Counsel also admitted that
he did not get in touch with the third witness (Dr. Roache)
until just before voir dire, and became aware of the final
witness (Martins) only partway through trial. Apart from
10 ANDRUS v. TEXAS
Per Curiam
some brief pretrial discussion with Dr. Roache, who averred
that he was “struck by the extent to which [counsel] ap-
peared unfamiliar” with pertinent issues, counsel did not
prepare the witnesses or go over their testimony before call-
ing them to the stand. 13 Habeas Tr., Def. Exh. 6, p. 3.
Over and over during the habeas hearing, counsel
acknowledged that he did not look into or present the myr-
iad tragic circumstances that marked Andrus’ life. For in-
stance, he did not know that Andrus had attempted suicide
in prison, or that Andrus’ experience in the custody of the
TYC left him badly traumatized. Aside from Andrus’
mother and biological father, counsel did not meet with any
of Andrus’ close family members, all of whom had disturb-
ing stories about Andrus’ upbringing. As a clinical psy-
chologist testified at the habeas hearing, Andrus suffered
“very pronounced trauma” and posttraumatic stress disor-
der symptoms from, among other things, “severe neglect”
and exposure to domestic violence, substance abuse, and
death in his childhood. 6 id., at 168–169, 180; 7 id., at 52.
Counsel uncovered none of that evidence. Instead, he
“abandoned [his] investigation of [Andrus’] background af-
ter having acquired only rudimentary knowledge of his his-
tory from a narrow set of sources.” Wiggins, 539 U. S., at
524.
On top of that, counsel “ignored pertinent avenues for in-
vestigation of which he should have been aware,” and in-
deed was aware. Porter, 558 U. S., at 40. At trial, counsel
averred that his review did not reveal that Andrus had any
mental-health issues. But materials prepared by a mitiga-
tion expert well before trial had pointed out that Andrus
had been “diagnosed with affective psychosis,” a mental-
health condition marked by symptoms such as depression,
mood lability, and emotional dysregulation. 3 id., at 70. At
the habeas hearing, counsel admitted that he “recall[ed]
noting,” based on the mitigation expert’s materials, that
Andrus had been “diagnosed with this seemingly serious
Cite as: 590 U. S. ____ (2020) 11
Per Curiam
mental health issue.” Id., at 71. He also acknowledged that
a clinical psychologist briefly retained to examine a limited
sample of Andrus’ files had informed him that Andrus may
have schizophrenia. Clearly, “the known evidence would
[have] le[d] a reasonable attorney to investigate further.”
Wiggins, 539 U. S., at 527. Yet counsel disregarded, rather
than explored, the multiple red flags.
In short, counsel performed virtually no investigation, ei-
ther of the few witnesses he called during the case in miti-
gation, or of the many circumstances in Andrus’ life that
could have served as powerful mitigating evidence. The un-
tapped body of mitigating evidence was, as the habeas hear-
ing revealed, simply vast.
“[C]ounsel’s failure to uncover and present [the] volumi-
nous mitigating evidence,” moreover, cannot “be justified as
a tactical decision.” Id., at 522; see also Williams, 529 U. S.,
at 396. Despite repeated questioning, counsel never of-
fered, and no evidence supports, any tactical rationale for
the pervasive oversights and lapses here. Instead, the over-
whelming weight of the record shows that counsel’s “failure
to investigate thoroughly resulted from inattention, not
reasoned strategic judgment.” Wiggins, 539 U. S., at 526.
That failure is all the more alarming given that counsel’s
purported strategy was to concede guilt and focus on miti-
gation. Indeed, counsel justified his decision to present “ba-
sically” “no defense” during the guilt phase by stressing
that he intended to train his efforts on the case in mitiga-
tion. 3 Habeas Tr. 57. As the habeas hearing laid bare,
that representation blinked reality. Simply put, “the scope
of counsel’s [mitigation] investigation” approached nonex-
istent. Wiggins, 539 U. S., at 528 (emphasis deleted).
2
No doubt due to counsel’s failure to investigate the case
in mitigation, much of the so-called mitigating evidence he
offered unwittingly aided the State’s case in aggravation.
12 ANDRUS v. TEXAS
Per Curiam
Counsel’s introduction of seemingly aggravating evidence
confirms the gaping distance between his performance at
trial and objectively reasonable professional judgment.
The testimony elicited from Andrus’ mother best illus-
trates this deficiency. First to testify during the case in mit-
igation, Andrus’ mother sketched a portrait of a tranquil
upbringing, during which Andrus got himself into trouble
despite his family’s best efforts. On her account, Andrus
fell into drugs entirely on his own: Drugs were not available
at home, Andrus did not use them at home, and she would
have intervened had she known about Andrus’ drug habits.
Andrus, his mother related to the jury, “[k]ind of ” “just de-
cided he didn’t want to do what [she] told him to do.” 49 Tr.
83.
Even though counsel called Andrus’ mother as a defense
witness, he was ill-prepared for her testimony. Andrus told
counsel that his mother was being untruthful on the stand,
but counsel made no real attempt to probe the accuracy of
her testimony. Later, at the habeas hearing, counsel con-
ceded that Andrus’ mother had been a “hostile” witness. 3
Habeas Tr. 94. He further admitted that he “[did not] know
if [Andrus’ mother] was telling the truth,” id., at 96, and
could not even say that he had known what Andrus’ mother
would say on the stand, because he had not “done any inde-
pendent investigation” of her, id., at 95.
None of that inaction was for want of warning. During
the habeas proceedings, a mitigation specialist averred that
she had alerted Andrus’ counsel to her concerns about An-
drus’ mother well before trial. In a short interview with the
mitigation specialist, Andrus’ mother had stated that she
“had too many kids,” and had taken out a $10,000 life-
insurance policy on Andrus on which she would be able to
collect were Andrus executed. 13 id., Def. Exh. 28, p. 5.
Troubled by these comments, the mitigation specialist “spe-
cifically discussed with [Andrus’ counsel] the fact that [An-
drus’ mother] was not being a cooperative witness and
Cite as: 590 U. S. ____ (2020) 13
Per Curiam
might not have Andrus’ best interests motivating her be-
havior.” Id., at 6. But Andrus’ counsel did not heed the
caution.
Turning a bad situation worse, counsel’s uninformed de-
cision to call Andrus’ mother ultimately undermined An-
drus’ own testimony. After Andrus testified that his mother
had sold drugs from home when he was a child, counsel
promptly pointed out that Andrus “heard [his] mama tes-
tify,” and that she “didn’t say anything about selling drugs.”
51 Tr. 48. Whether counsel merely intended to provide An-
drus an opportunity to explain the discrepancy (or, far
worse, sought to signal that his client was being deceitful)
the jury could have understood counsel’s statements to
insinuate that Andrus was lying. Counsel did nothing to
dislodge that suggestion, and the damaging exchange oc-
curred only because defense counsel had called a hostile
witness in the first place. Plainly, these offerings of seem-
ingly aggravating evidence further demonstrate counsel’s
constitutionally deficient performance.
3
Counsel also failed to conduct any independent investiga-
tion of the State’s case in aggravation, despite ample oppor-
tunity to do so. He thus could not, and did not, rebut critical
aggravating evidence. This failure, too, reinforces counsel’s
deficient performance. See Rompilla v. Beard, 545 U. S.
374, 385 (2005) (“counsel ha[s] a duty to make all reasona-
ble efforts to learn what they c[an] about the offense[s]” the
prosecution intends to present as aggravating evidence).
During the case in aggravation, the State’s task was to
prove to the jury that Andrus presented a future danger to
society. Tex. Code Crim. Proc. Ann., Art. 37.071, §2(b)(1)
(Vernon 2006). To that end, the State emphasized that An-
drus had acted aggressively in TYC facilities and in prison
while awaiting trial. This evidence principally comprised
14 ANDRUS v. TEXAS
Per Curiam
verbal threats, but also included instances of Andrus’ kick-
ing, hitting, and throwing excrement at prison officials
when they tried to control him. See App. to Pet. for Cert.
10–13. Had counsel genuinely investigated Andrus’ experi-
ences in TYC custody, counsel would have learned that An-
drus’ behavioral problems there were notably mild, and the
harms he sustained severe.2 Or, with sufficient under-
standing of the violent environments Andrus inhabited his
entire life, counsel could have provided a counternarrative
of Andrus’ later episodes in prison. But instead, counsel left
all of that aggravating evidence untouched at trial—even
going so far as to inform the jury that the evidence made it
“probabl[e]” that Andrus was “a violent kind of guy.” 52 Tr.
35.
The State’s case in aggravation also highlighted Andrus’
alleged commission of a knifepoint robbery at a dry-
cleaning business. At the time of the offense, “all [that] the
crime victim . . . told the police . . . was that he had been the
victim of an assault by a black man.” 3 Habeas Tr. 65. Alt-
hough Andrus stressed to counsel his innocence of the of-
fense, and although the State had not proceeded with
charges, Andrus’ counsel did not attempt to exclude or re-
but the State’s evidence. That, too, is because Andrus’
counsel concededly had not independently investigated the
incident. In fact, at the habeas hearing, counsel did not
even recall Andrus’ denying responsibility for the offense.
——————
2 See, e.g., 5 Habeas Tr. 189 (TYC ombudsman testifying that it was
“surpris[ing] how few” citations Andrus received, “particularly in the
dorms where [Andrus] was” housed); ibid. (TYC ombudsman finding
“nothing uncommon” about Andrus’ altercations because “sometimes you
. . . have to fight to get by” in the “violent atmosphere” and “savage envi-
ronment”); id., at 169 (TYC ombudsman testifying that Andrus’ isolation
periods in TCY custody, for 90 days at a time when Andrus was 16 or 17
years old, “would horrify most current professionals in our justice field
today”); id., at 246 (TYC ombudsman testifying that Andrus’ “experience
at TYC” “damaged him” and “further traumatized” him).
Cite as: 590 U. S. ____ (2020) 15
Per Curiam
Had he looked, counsel would have discovered that the only
evidence originally tying Andrus to the incident was a lone
witness statement, later recanted by the witness,3 that led
to the inclusion of Andrus’ photograph in a belated photo
array, which the police admitted gave rise to numerous re-
liability concerns. The dissent thus reinforces Andrus’
claim of deficient performance by recounting and emphasiz-
ing the details of the dry-cleaning offense as if Andrus were
undoubtedly the perpetrator. See post, at 6 (opinion of
ALITO, J.). The very problem here is that the jury indeed
heard that account, but not any of the significant evidence
that would have cast doubt on Andrus’ involvement in the
offense at all: significant evidence that counsel concededly
failed to investigate.4
That is hardly the work of reasonable counsel. In Texas,
a jury cannot recommend a death sentence without unani-
mously finding that a defendant presents a future danger
to society (i.e., that the State has made a sufficient showing
of aggravation). Tex. Code Crim. Proc. Ann., Art. 37.071,
——————
3 The dissent maintains that this witness, Andrus’ ex-girlfriend,
“linked [Andrus] to the robbery,” post, at 6, n. 4, even though she testified
at the habeas hearing that she thought “it was impossible” that Andrus
had committed the offense, 8 Habeas Tr. 57.
4 The dissent does not mention that Andrus’ image was conspicuously
placed in a central position in the photo array, as the “[o]nly one . . . look-
ing directly up and out.” 8 Habeas Tr. 35; see also id., at 32. Nor does
the dissent acknowledge that there was an approximately 3-month in-
terval between the incident (after which the victim provided little iden-
tifying information about the assailant) and the police’s presentation of
the photo array to the victim. See id., at 37; 46 Tr. 65. When asked about
the delay, the detective who prepared the photo array admitted that
memory can “deca[y] within a matter of days after a traumatizing inci-
dent like a crime” and that an “eyewitness identificatio[n]” “can be”
“more exponentially problematic” “the greater the time interval between
the incident and the identification.” 8 Habeas Tr. 31; see also ibid. (de-
tective confirming that there can be “real problems with reliability” if an
“identification [was] made several months” after).
16 ANDRUS v. TEXAS
Per Curiam
§2(b)(1). Only after a jury makes a finding of future dan-
gerousness can it consider any mitigating evidence. Ibid.
Thus, by failing to conduct even a marginally adequate in-
vestigation, counsel not only “seriously compromis[ed his]
opportunity to respond to a case for aggravation,” Rompilla,
545 U. S., at 385, but also relinquished the first of only two
procedural pathways for opposing the State’s pursuit of the
death penalty. There is no squaring that conduct, certainly
when examined alongside counsel’s other shortfalls, with
objectively reasonable judgment.
B
Having found deficient performance, the question re-
mains whether counsel’s deficient performance prejudiced
Andrus. See Strickland, 466 U. S., at 692. Here, prejudice
exists if there is a reasonable probability that, but for his
counsel’s ineffectiveness, the jury would have made a dif-
ferent judgment about whether Andrus deserved the death
penalty as opposed to a lesser sentence. See Wiggins, 539
U. S., at 536; see also Tex. Code Crim. Proc. Ann., Art.
37.071, §2(e)(1). In assessing whether Andrus has made
that showing, the reviewing court must consider “the total-
ity of the available mitigation evidence—both that adduced
at trial, and the evidence adduced in the habeas proceed-
ing”—and “reweig[h] it against the evidence in aggrava-
tion.” Williams, 529 U. S., at 397–398; see also Sears v. Up-
ton, 561 U. S. 945, 956 (2010) (per curiam) (“A proper
analysis of prejudice under Strickland would have taken
into account the newly uncovered [mitigation] evidence . . . ,
along with the mitigation evidence introduced during [the
defendant’s] penalty phase trial, to assess whether there is
a reasonable probability that [the defendant] would have
received a different sentence after a constitutionally suffi-
cient mitigation investigation” (citing cases)). And because
Andrus’ death sentence required a unanimous jury recom-
Cite as: 590 U. S. ____ (2020) 17
Per Curiam
mendation, Tex. Code Crim. Proc. Ann., Art. 37.071, preju-
dice here requires only “a reasonable probability that at
least one juror would have struck a different balance” re-
garding Andrus’ “moral culpability,” Wiggins, 539 U. S., at
537–538; see also Tex. Code Crim. Proc. Ann., Art. 37.071,
§2(e)(1).
According to Andrus, effective counsel would have
painted a vividly different tableau of aggravating and miti-
gating evidence than that presented at trial. See Pet. for
Cert. 18. But despite powerful and readily available miti-
gating evidence, Andrus argues, the Texas Court of Crimi-
nal Appeals failed to engage in any meaningful prejudice
inquiry. See ibid.
It is unclear whether the Court of Criminal Appeals con-
sidered Strickland prejudice at all. Its one-sentence denial
of Andrus’ Strickland claim, see supra, at 7, does not con-
clusively reveal whether it determined that Andrus had
failed to demonstrate deficient performance under Strick-
land’s first prong, that Andrus had failed to demonstrate
prejudice under Strickland’s second prong, or that Andrus
had failed to satisfy both prongs of Strickland.
Unlike the concurring opinion, however, the brief order of
the Court of Criminal Appeals did not analyze Strickland
prejudice or engage with the effect the additional mitigat-
ing evidence highlighted by Andrus would have had on the
jury.5 What little is evident from the proceeding below is
——————
5 The Court of Criminal Appeals did briefly observe that the trial
court’s order recommending relief had omitted the “ ‘reasonable probabili-
ty’ ” language when reciting the Strickland prejudice standard. App. to
Pet. for Cert. 8, n. 2; cf. Strickland, 466 U. S., at 694 (a defendant “must
show that there is a reasonable probability that, but for counsel’s unpro-
fessional errors, the result of the proceeding would have been different”).
Even were there reason to set aside that “[t]rial judges are presumed to
know the law,” Lambrix v. Singletary, 520 U. S. 518, 532, n. 4 (1997) (in-
ternal quotation marks omitted), the trial court’s omission of the “rea-
sonable probability” language would at most suggest that it held Andrus
18 ANDRUS v. TEXAS
Per Curiam
that the concurring opinion’s analysis of or conclusion re-
garding prejudice did not garner a majority of the Court of
Criminal Appeals.6 Given that, the court may have con-
cluded simply that Andrus failed to demonstrate deficient
performance under the first prong of Strickland (without
even reaching the second prong). For the reasons explained
above, any such conclusion is erroneous as a matter of law.
See supra, at 8–16.
The record before us raises a significant question whether
the apparent “tidal wave,” 7 Habeas Tr. 101, of “available
mitigating evidence taken as a whole” might have suffi-
ciently “ ‘influenced the jury’s appraisal’ of [Andrus’] moral
culpability” as to establish Strickland prejudice, Wiggins,
539 U. S., at 538 (quoting Williams, 529 U. S., at 398).
(That is, at the very least, whether there is a reasonable
probability that “at least one juror would have struck a dif-
ferent balance.” Wiggins, 539 U. S., at 537.) That prejudice
inquiry “necessarily require[s] a court to ‘speculate’ as to
the effect of the new evidence” on the trial evidence, “re-
gardless of how much or little mitigation evidence was pre-
sented during the initial penalty phase.” Sears, 561 U. S.,
at 956; see also id., at 954 (“We have never limited the prej-
udice inquiry under Strickland to cases in which there was
——————
to (and found that Andrus had satisfied) a stricter standard of prejudice
than that set forth in Strickland.
6 The concurring opinion, moreover, seemed to assume that the preju-
dice inquiry here turns principally on how the facts of this case compare
to the facts in Wiggins. We note that we have never before equated what
was sufficient in Wiggins with what is necessary to establish prejudice.
Cf. Wiggins, 539 U. S., at 537–538 (“[T]he mitigating evidence in this
case is stronger, and the State’s evidence in support of the death penalty
far weaker, than in Williams, where we found prejudice as the result of
counsel’s failure to investigate and present mitigating evidence”); Wil-
liams, 529 U. S., at 399 (finding such prejudice after applying AEDPA
deference).
Cite as: 590 U. S. ____ (2020) 19
Per Curiam
‘little or no mitigation evidence’ presented”).7 Given the un-
certainty as to whether the Texas Court of Criminal Ap-
peals adequately conducted that weighty and record-in-
tensive analysis in the first instance, we remand for the
Court of Criminal Appeals to address Strickland prejudice
in light of the correct legal principles articulated above. See
Cutter v. Wilkinson, 544 U. S. 709, 718, n. 7 (2005).
* * *
We conclude that Andrus has shown deficient perfor-
mance under the first prong of Strickland, and that there is
a significant question whether the Court of Criminal Ap-
peals properly considered prejudice under the second prong
of Strickland. We thus grant Andrus’ petition for a writ of
certiorari and his motion for leave to proceed in forma pau-
peris, vacate the judgment of the Texas Court of Criminal
Appeals, and remand the case for the court to address the
prejudice prong of Strickland in a manner not inconsistent
with this opinion.
It is so ordered.
——————
7 The dissent trains its attention on the aggravating evidence actually
presented at trial. Post, at 4–7; but see Sears, 561 U. S., at 956 (Strick-
land prejudice inquiry “will necessarily require a court to ‘speculate’ as
to the effect of the new evidence” on the trial evidence); 561 U. S., at 956
(“A proper analysis of prejudice under Strickland would have taken into
account the newly uncovered evidence . . . , along with the mitigation ev-
idence introduced during [the] penalty phase trial”).
Cite as: 590 U. S. ____ (2020) 1
ALITO, J., dissenting
SUPREME COURT OF THE UNITED STATES
TERENCE TRAMAINE ANDRUS v. TEXAS
ON PETITION FOR WRIT OF CERTIORARI TO THE
COURT OF CRIMINAL APPEALS OF TEXAS
No. 18–9674. Decided June 15, 2020
JUSTICE ALITO, with whom JUSTICE THOMAS and
JUSTICE GORSUCH join, dissenting.
The Court clears this case off the docket, but it does so on
a ground that is hard to take seriously. According to the
Court, “[i]t is unclear whether the Court of Criminal Ap-
peals considered Strickland prejudice at all.” Ante, at 17;
see Strickland v. Washington, 466 U. S. 668 (1984). But
that reading is squarely contradicted by the opinion of the
Court of Criminal Appeals (CCA), which said explicitly that
Andrus failed to show prejudice:
“[Andrus] fails to meet his burden under Strickland v.
Washington, 466 U. S. 668 (1984), to show by a prepon-
derance of the evidence that his counsel’s representa-
tion fell below an objective standard of reasonableness
and that there was a reasonable probability that the re-
sult of the proceedings would have been different, but
for counsel’s deficient performance.” App. to Pet. for
Cert. 7–8 (emphasis added).
Not only does the CCA opinion contain this express state-
ment, but it adds that the trial court did not heed Strick-
land’s test for prejudice. See App. to Pet. for Cert. 8, n. 2
(“[T]hroughout its findings, the trial court misstates the
Strickland prejudice standard by omitting the standard’s
‘reasonable probability’ language”). And the record clearly
shows that the trial court did not apply that test to Andrus’s
claim. See App. to Pet. for Cert. 36–37. A majority of this
Court cannot seriously think that the CCA pointed this out
and then declined to reach the issue of prejudice.
2 ANDRUS v. TEXAS
ALITO, J., dissenting
How, then, can the Court get around the unmistakable
evidence that the CCA decided the issue of prejudice? It
begins by expressing doubt about the meaning of the critical
sentence reproduced above. According to the Court, that
sentence “does not conclusively reveal whether [the CCA]
determined . . . that Andrus had failed to demonstrate prej-
udice under Strickland’s second prong.” Ante, at 17. It is
hard to write a more conclusive sentence than “[Andrus]
fails to meet his burden under Strickland v. Washington,
466 U. S. 668 (1984), to show by a preponderance of the ev-
idence . . . that there was a reasonable probability that the
result of the proceedings would have been different, but for
counsel’s deficient performance.” App. to Pet. for Cert. 7–8.
Perhaps the Court thinks the CCA should have used
CAPITAL LETTERS or bold type. Or maybe it should
have added: “And we really mean it!!!”
Not only does the Court express doubt that the CCA
reached the prejudice prong of Strickland, but the Court is
not sure that the CCA decided even the performance prong.
See ante, at 17 (“Its one-sentence denial of Andrus’ Strick-
land claim . . . does not conclusively reveal whether it de-
termined that Andrus had failed to demonstrate deficient
performance under Strickland’s first prong”). The Court
may feel it necessary to make that statement because the
CCA disposed of both prongs in the sentence quoted above.
So if that sentence is not sufficient to show that the CCA
reached the prejudice prong, there is no better reason for
thinking that it decided the performance prong. But if the
Court really thinks that the CCA did not decide the perfor-
mance issue, why does it treat that issue differently from
the prejudice issue? Why does it decide the performance
question in the first instance? Are we now a court of “first
view” and not, as we have often stressed, a “court of re-
view”? See, e.g., McLane Co., Inc. v. EEOC, 581 U. S. ___,
___ (2017) (slip op., at 11). The Court’s disparate treatment
of the two parts of the CCA’s dispositive sentence shows
Cite as: 590 U. S. ____ (2020) 3
ALITO, J., dissenting
that the Court is only selectively skeptical.
The Court gives two reasons for doubting that the CCA
reached the issue of prejudice, but both are patent make-
weights. First, the Court notes that the CCA’s per curiam
opinion, unlike the concurring opinion, did not provide rea-
sons for finding that prejudice had not been shown. But the
failure to explain is not the same as failure to decide. To-
day’s “tutelary remand” is a misuse of our supervisory au-
thority and a waste of our and the CCA’s time. Lawrence v.
Chater, 516 U. S. 163, 185 (1996) (Scalia, J., dissenting).
Second, the Court observes that the concurring opinion,
which discussed the question of prejudice at some length,
was joined by only four of the CCA’s nine judges. See App.
to Pet. for Cert. 9–21 (opinion of Richardson, J., joined by
Keller, P. J., and Hervey and Slaughter, JJ.). But that does
not show that the other five declined to decide the question
of prejudice. The most that one might possibly infer is that
these judges might not have agreed with everything in the
concurrence, but even that is by no means a certainty. So
the Court’s reading of the decision below is contrary to the
plain language of the decision and is not supported by any
reason worth mentioning.
If that were not enough, the Court’s reading is belied by
Andrus’s interpretation of the CCA decision. Andrus no-
where claims that the CCA failed to decide the issue of prej-
udice. On the contrary, the petition faults the CCA for
providing “a truncated ‘no prejudice’ analysis,” not for fail-
ing to decide the prejudice issue at all. Pet. for Cert. ii (em-
phasis added). Indeed, the main argument in the petition
is that we should modify Strickland because courts are too
often rejecting ineffective-assistance claims for lack of prej-
udice. That argument would make no sense if the CCA had
not decided the prejudice issue, something that is never
even implied by Andrus’s counsel in either the 40-page pe-
tition or the 11-page reply.
Not only did the CCA clearly hold that Andrus failed to
4 ANDRUS v. TEXAS
ALITO, J., dissenting
show prejudice, but there was strong support for that hold-
ing in the record. To establish prejudice, Andrus must show
“a substantial, not just conceivable, likelihood” that one of
the jurors who unanimously agreed on his sentence would
not have done so if his trial counsel had presented more
mitigation evidence. Cullen v. Pinholster, 563 U. S. 170,
189 (2011) (internal quotation marks omitted). This in-
quiry focuses not just on the newly offered mitigation evi-
dence, but on the likelihood that this evidence would have
overcome the State’s aggravation evidence. See, e.g., Sears
v. Upton, 561 U. S. 945, 955–956 (2010) (per curiam). While
providing a lengthy (and one-sided) discussion of Andrus’s
mitigation evidence, the Court never acknowledges the vol-
ume of evidence that Andrus is prone to brutal and sense-
less violence and presents a serious danger to those he en-
counters whether in or out of prison. Instead, the Court
says as little as possible about Andrus’s violent record.
For example, here is what the Court says about the
crimes for which he was sentenced to death: “Not long after
Andrus’ release from prison at age 18, Andrus attempted
the fatal carjacking that resulted in his capital convictions.”
Ante, at 6.
Here is what the record shows. According to Andrus’s
confession, he left his apartment one evening, “ ‘amped up’
on embalming fluid [PCP] mixed with marijuana, cocaine,
and beer,” and looked for a car to “go joy-riding.” No. AP–
76,936, p. 5 (CCA, Mar. 23, 2016) (Reh’g Op.); see also 54
Tr., Pl. Exh. 147 (Andrus’s confession). In the parking lot
of a supermarket, he saw Avelino Diaz drop off his wife,
Patty, in front of the store. By his own admission, Andrus
approached Diaz’s car with a gun drawn, but he abandoned
the carjacking attempt when he saw that the car had a stick
shift, which he could not drive. Alerted by a store employee,
Patty Diaz ran out of the store and found her husband lying
by the side of the car with a bullet wound in the back of his
head. He was subsequently pronounced dead.
Cite as: 590 U. S. ____ (2020) 5
ALITO, J., dissenting
After killing Avelino Diaz, Andrus approached a car with
two occupants, whom Andrus described as an “old man and
old wom[a]n.” Id., at 2. Andrus fired three shots into the
car. The first went through the open driver’s side window
and hit the passenger, Kim-Phuong Vu Bui, in the head. As
the car sped away, Andrus fired a second shot, which en-
tered the back driver’s side window, and a third shot, which
“entered at an angle indicating that the shot originated
from a farther distance.” Reh’g Op. 3. One of these bullets
hit the driver, Steve Bui, in the back. Seeing that blood was
coming out of his wife’s mouth, Steve drove her to a hospital
and carried her inside, where she died.
These senseless murders in October 2008 were not An-
drus’s first crimes. In 2004, he was placed on probation for
a drug offense, but just two weeks later, he committed an
armed robbery. Andrus and two others followed a woman
to her parents’ home, where they held her at gunpoint and
took her purse and gym bag. The woman identified Andrus
as the perpetrator who held the gun. Id., at 7.1
For this offense, Andrus was sent to a juvenile facility
where he showed such “ ‘significant assaultive behavior’ to-
ward other youths and staff ” that he was eventually trans-
ferred to an adult facility. App. to Pet. for Cert. 11.2 Shortly
——————
1 The Court credits Andrus’s version of the event and repeats his alle-
gation that he merely served as a “lookout.” Ante, at 1, 6. As the CCA
explained on direct review, however, the victim matched Andrus’s cloth-
ing to the gunman’s. See Reh’g Op. 7; see also 46 Tr. 23–25 (arresting
officer explaining that only Andrus’s clothing matched the suspect de-
scription).
2 Just as the Court provides a one-sided summary of Andrus’s mitiga-
tion evidence, it quibbles at every possible turn with the aggravation ev-
idence. Thus, the Court states that Andrus’s behavioral problems at this
facility “were notably mild.” Ante, at 14. But the witness on whose tes-
timony the Court relies admitted that Andrus’s record included multiple
threats and assaults against staff and other youths. 4 Habeas Tr. 202–
204. And the record shows that Andrus had needed to be removed from
general population 77 times. 10 id., Pl. Exh. 28. The responsible correc-
6 ANDRUS v. TEXAS
ALITO, J., dissenting
after his release, he again violated his supervisory condi-
tions and was returned to the adult facility. Ibid.
When he was released again, he committed an armed rob-
bery of a dry-cleaning establishment. Around 7 a.m. one
morning, he entered the business and chased the owner,
Tuan Tran, to the back. He beat Tran and threatened him
with a knife until Tran gave him money. Reh’g Op. 7–8.
Andrus’s ex-girlfriend told the police that he confessed to
this robbery. 8 Habeas Tr. 14.3 In addition, Tran picked
Andrus out of a photo array, 46 Tr. 66, 69–70,4 and testified
at trial that the robber was in the courtroom, id., at 59–60,
but he was too afraid to point at Andrus, ibid. Less than
two months after this crime, Andrus murdered Avelino
Diaz and Kim-Phuong Vu Bui. App. to Pet. for Cert. 11.
While awaiting trial for those murders, Andrus carried
out a reign of terror in jail. He assaulted another detainee,
attacked and injured corrections officers, threw urine in an
officer’s face, repeatedly made explicit threats to kill offic-
ers and staff, flooded his cell and threw excrement on the
——————
tions officials obviously did not think this record was “notably mild,” be-
cause it prompted them to transfer him to an adult facility.
3 Although Andrus’s ex-girlfriend later signed an affidavit contradict-
ing herself, 41 id., Def. Exh. 139, pp. 1–2, she admitted at the habeas
hearing—after learning that she had been recorded—that she indeed re-
layed this information, 8 id., at 48–49. Andrus’s counsel tried to with-
draw her affidavit from evidence, having “learned information that
caused [them] to doubt [her] reliability.” Id., at 5.
4 The Court again credits Andrus’s allegation that he did not commit
this robbery. See ante, at 14–15. In support, the Court points to what
Tran told police shortly after being beaten and to supposed problems
with the photo array from which Tran first identified Andrus. But the
Court cannot dispute that Andrus’s ex-girlfriend linked him to the rob-
bery or that Tran identified him twice. Nor did the detective to whom
the Court refers in fact testify that “the inclusion of Andrus’ photograph
in a belated photo array . . . gave rise to numerous reliability concerns.”
Ante, at 15; see 8 Habeas Tr. 31 (testifying, in response to habeas coun-
sel’s repeated questions whether delays affect the reliability of identifi-
cations, only that they “can”); id., at 42–44 (affirming the bases for An-
drus’s inclusion).
Cite as: 590 U. S. ____ (2020) 7
ALITO, J., dissenting
walls, and engaged in other disruptive acts. Id., at 11–13.
Also while awaiting trial for murder, he had the words
“murder weapon” tattooed on his hands and a smoking gun
tattooed on his forearm. 51 Tr. 65–66, 68.
In sum, the CCA assessed the issue of prejudice in light
of more than the potentially mitigating evidence that the
Court marshals for Andrus. The CCA had before it strong
aggravating evidence that Andrus wantonly killed two in-
nocent victims and shot a third; that he committed other
violent crimes; that he has a violent, dangerous, and unsta-
ble character; and that he is a threat to those he encounters.
The CCA has already held once that Andrus failed to
establish prejudice. I see no good reason why it should be
required to revisit the issue.