FILED
United States Court of Appeals
PUBLISH Tenth Circuit
UNITED STATES COURT OF APPEALS October 28, 2019
Elisabeth A. Shumaker
FOR THE TENTH CIRCUIT Clerk of Court
_________________________________
JIMMY DEAN HARRIS,
Petitioner - Appellant,
v. No. 17-6109
TOMMY SHARP, Interim Warden,
Oklahoma State Penitentiary, *
Respondent - Appellee.
_________________________________
Appeal from the United States District Court
for the Western District of Oklahoma
(D.C. No. 5:08-CV-00375-F)
_________________________________
Jack Fisher, Fisher Law Office, Edmond, Oklahoma, and Emma V. Rolls,
Assistant Federal Public Defender, Oklahoma City, Oklahoma, on behalf of
the Petitioner-Appellant.
Jennifer L. Crabb, Assistant Attorney General (Mike Hunter, Attorney
General of Oklahoma, with her on the briefs), Oklahoma City, Oklahoma,
on behalf of the Respondent-Appellee.
_________________________________
Before TYMKOVICH, Chief Judge, BACHARACH, and McHUGH,
Circuit Judges.
_________________________________
BACHARACH, Circuit Judge.
_________________________________
*
Pursuant to Fed. R. App. P. 42(c)(2), Mike Carpenter is replaced by
Tommy Sharp, as the Interim Warden of the Oklahoma State Penitentiary.
TABLE OF CONTENTS
Background ................................................................................... 2
The Standard of Review ................................................................. 3
Appellate Arguments Covered in an
Existing Certificate of Appealability .............................................. 5
I. Ineffective Assistance of Counsel ........................................... 5
A. The Strickland Standard ................................................ 6
B. Failure to Seek a Pretrial Hearing on Intellectual
Disability as a Bar to Execution .................................... 7
1. The Standard of Review ........................................ 9
2. Deficiency Prong .................................................. 13
3. Prejudice Prong .................................................... 18
(a) Unreasonable Determination of Fact .............. 19
(b) The Need for an Evidentiary Hearing ............. 26
(c) Conclusion ................................................... 36
C. Failure to Adequately Present Mitigation Evidence .......... 37
1. The Legal Standard and the Standard of Review ...... 38
2. Intellectual Impairment as a Mitigating Factor ........ 38
(a) Evidence of an Intellectual Impairment .......... 39
(b) Mitigation Evidence Involving an Intellectual
Disability .................................................... 41
(c) Mitigation Evidence Involving Borderline
Intellectual Functioning ................................ 45
i
i. The OCCA’s Reliance on Both Prongs
(Deficient Performance and Prejudice) .. 45
ii. Deficient Performance .......................... 46
iii. Prejudice ............................................. 50
(d) Mitigation Evidence Involving Mental
Illness ......................................................... 52
i. Mental Health Evidence in the
2005 Retrial ........................................ 53
ii. Other Existing Evidence of Mr. Harris’s
Mental Illness ...................................... 54
iii. Claim of Ineffective Assistance of
Counsel ............................................... 55
a. Deficiency Prong ......................... 56
(i) Unreasonable Factual
Determinations ................... 56
(ii) Unreasonable Application of
Supreme Court Precedents ... 58
b. Prejudice .................................... 59
II. Jury Instructions and Closing Arguments as to
Mitigation Evidence ............................................................... 63
A. The Standard of Review ................................................. 64
B. The Jury Instruction ...................................................... 66
C. The Prosecutors’ Closing Arguments .............................. 67
1. Applicability of 28 U.S.C. § 2254(d) ...................... 69
2. Unreasonable Determination of Fact ....................... 70
3. Unreasonable Application of Supreme Court
Precedent ............................................................. 74
ii
III. Victim-Impact Testimony ....................................................... 81
A. The Constitutional Limit on Victim-Impact Testimony ..... 81
B. The Victim-Impact Testimony and the Issue of
Harmlessness ................................................................ 82
C. Structural or Harmless Error .......................................... 83
D. Harmlessness ................................................................ 85
IV. Cumulative Error ................................................................... 89
Motion to Expand the Certificate of Appealability ............................ 91
Conclusion .................................................................................... 94
iii
Mr. Jimmy Dean Harris was convicted of first-degree murder and
sentenced to death. He appealed, and the Oklahoma Court of Criminal
Appeals (OCCA) reversed his sentence and remanded for a retrial at the
penalty phase. After the retrial, the state district court reimposed the death
penalty. Mr. Harris appealed and sought post-conviction relief in state
court. When these efforts failed, he brought a habeas petition in federal
district court. The court denied relief, and Mr. Harris appeals.
On appeal, Mr. Harris argues in part that his trial counsel was
ineffective in failing to seek a pretrial hearing on the existence of an
intellectual disability, which would have prevented the death penalty. 1 The
federal district court rejected this claim. In our view, the district court
should have conducted an evidentiary hearing to decide this claim, so we
reverse and remand for further consideration. Given the need to remand on
this issue, we also remand for the district court to reconsider the claim of
cumulative error. But we affirm the denial of habeas relief on Mr. Harris’s
other claims.
1
Older opinions often used the term “mentally retarded.” See, e.g.,
Atkins v. Virginia, 536 U.S. 304, 316 (2002). But more recently, we have
used the term “intellectually disabled.” See Postelle v. Carpenter, 901 F.3d
1202, 1210 n.4 (10th Cir. 2018); cf. Rosa’s Law, Pub. L. No. 111-256, 124
Stat. 2643 (2010) (changing references in federal law from “mental
retardation” and “mentally retarded” to “intellectual disability” and
“intellectually disabled”).
1
Background 2
Jimmy Dean Harris and Pam Harris were married for about twenty
years. Mr. Harris repaired transmissions, as did Pam, who worked for Mr.
Merle Taylor. With the passage of time came marital strain between Mr.
Harris and Pam.
In 1999, Pam obtained a divorce and restraining order, requiring Mr.
Harris to move out of their house. He complied, moving his belongings
into a storage shed, but he grew distraught—crying, drinking, and taking
Valium.
The next day, Pam returned home and discovered that Mr. Harris had
vandalized the house and moved some of her belongings into the storage
shed. This incident led Pam to change the locks and to obtain a second
restraining order, which required Mr. Harris to stay away from the house.
Mr. Harris repeatedly asked Pam to allow him to retrieve his tools.
After a few days, Mr. Harris went to Pam’s workplace and shot at her, Mr.
Taylor, and his daughter (Jennifer Taylor). Mr. Taylor died, Pam was
wounded, and Jennifer Taylor escaped without injury.
2
Under the Antiterrorism and Effective Death Penalty Act (AEDPA),
we defer to the OCCA’s factual findings absent clear and convincing
evidence to the contrary. 28 U.S.C. § 2254(e)(1). We thus state the facts as
the OCCA found them unless noted otherwise.
2
At a 2001 trial, the jury found Mr. Harris guilty of first-degree
murder in the death of Merle Taylor and recommended the death penalty,
finding one aggravating circumstance (creation of a substantial risk of
death to more than one person). 3 As noted above, the death sentence was
vacated by the OCCA in a prior appeal. At the 2005 retrial on the penalty,
the prosecution alleged two aggravating factors:
1. Mr. Harris created a substantial risk of death to more than one
person.
2. Mr. Harris posed a continuing threat to society.
The jury found both aggravating factors and again recommended the death
penalty. The trial court agreed with the recommendation and resentenced
Mr. Harris to the death penalty.
The Standard of Review
We engage in de novo review of the federal district court’s legal
analysis. Littlejohn v. Trammell, 704 F.3d 817, 825 (10th Cir. 2013). In
district court, review is deferential when the state appellate court rejects a
claim on the merits. After rejection of the claim in state court, the federal
district court can reach the merits only if the state appellate court’s
decision was
contrary to, or involving an unreasonable application of,
clearly established Federal law, as determined by the Supreme
Court of the United States; or
3
The jury also found Mr. Harris guilty of attempted murder as to Pam.
3
based on an unreasonable determination of the facts given the
evidence presented in state court.
Antiterrorism and Effective Death Penalty Act of 1996 (AEDPA), 28
U.S.C. § 2254(d).
To determine whether a state-court decision was contrary to, or
involved an unreasonable application of, clearly established law, we
engage in a two-step process. Budder v. Addison, 851 F.3d 1047, 1051
(10th Cir.), cert. denied, 138 S. Ct. 475 (2017). We first determine the
clearly established law by considering Supreme Court precedent. Williams
v. Taylor, 529 U.S. 362, 379 (2000). We then ask whether the state court’s
decision was contrary to, or involved an unreasonable application of, that
precedent. Id.
We must defer to the state court’s factual findings unless “the state
court[] plainly misapprehend[ed] or misstate[d] the record in making [its]
findings, and the misapprehension goes to a material factual issue that is
central to [the] petitioner’s claim.” Ryder ex rel. Ryder v. Warrior, 810
F.3d 724, 739 (10th Cir. 2016) (quoting Byrd v. Workman, 645 F.3d 1159,
1171–72 (10th Cir. 2011)). To overcome the state appellate court’s factual
findings, the petitioner must show that they are objectively unreasonable.
Smith v. Aldridge, 904 F.3d 874, 880 (10th Cir. 2018).
If the state’s highest court acted unreasonably in applying Supreme
Court precedent or finding facts, the district court must decide whether the
4
conviction or sentence violated the Constitution. See Fry v. Pliler, 551
U.S. 112, 119 (2007) (stating that 28 U.S.C. § 2254(d) provides
“precondition[s] to the grant of habeas relief . . . , not an entitlement to
it”); Hancock v. Trammell, 798 F.3d 1002, 1010 (10th Cir. 2015) (“[E]ven
when petitioners satisfy the threshold in § 2254(d), they must establish a
violation of federal law or the federal constitution.”).
Appellate Arguments Covered in an
Existing Certificate of Appealability
Our court previously granted a certificate of appealability on Mr.
Harris’s appellate arguments involving ineffective assistance of counsel,
an improper jury instruction on mitigation evidence, improper closing
arguments about the mitigation evidence, improper victim testimony
recommending a particular sentence, and cumulative error. We reverse and
remand for further consideration of the claims involving (1) ineffective
assistance in the failure to seek a pretrial hearing on an intellectual
disability and (2) cumulative error.
I. Ineffective Assistance of Counsel
The Sixth Amendment entitles a defendant to effective assistance of
counsel. Strickland v. Washington, 466 U.S. 668 (1984). Invoking this
amendment, Mr. Harris argues that his attorney at the 2005 retrial was
ineffective for failing to
seek a pretrial hearing on the existence of an intellectual
disability, which would have precluded the death penalty,
5
present additional trial evidence for mitigation based on an
intellectual disability, and
present additional mitigation evidence at trial regarding a
lesser intellectual impairment or mental illness.
A. The Strickland Standard
To address Mr. Harris’s arguments, the district court needed to apply
the two-part test set out in Strickland v. Washington, 466 U.S. 668 (1984).
Under the first part of the test, the court was to determine whether
Mr. Harris’s attorney was deficient. Attorneys are deficient when their
mistakes are so serious that they stop functioning as “counsel” for
purposes of the Sixth Amendment. Id. at 687. In making this determination,
the court ordinarily presumes that counsel’s performance is reasonable and
might entail a sound strategy. Newmiller v. Raemisch, 877 F.3d 1178, 1196
(10th Cir. 2017). In capital cases, however, courts scrutinize attorney
performance particularly closely in the sentencing phase. Littlejohn v.
Trammel, 704 F.3d 817, 859 (10th Cir. 2013).
To overcome this presumption, a petitioner “must show that
counsel’s representation fell below an objective standard of
reasonableness.” Strickland, 466 U.S. at 688. This inquiry is “highly
deferential,” and courts should avoid “the distorting effects of hindsight.”
Id. at 689. Strategic decisions after a “thorough investigation” are afforded
even greater deference and are “virtually unchallengeable.” Id. at 690.
6
“Even under de novo review, the standard for judging counsel’s
representation is a most deferential one.” Harrington v. Richter, 562 U.S.
86, 105 (2011).
When a habeas petitioner alleges ineffective assistance of counsel,
deference exists both in the underlying constitutional test (Strickland) and
the AEDPA’s standard for habeas relief, creating a “doubly deferential
judicial review.” Knowles v. Mirzayance, 556 U.S. 111, 123 (2009). Under
this double deference, we consider “whether there is any reasonable
argument that counsel satisfied Strickland’s deferential standard.” Ellis v.
Raemisch, 872 F.3d 1064, 1084 (10th Cir. 2017) (quoting Harrington v.
Richter, 562 U.S. 86, 105 (2011) (emphasis in original)).
The petitioner must show not only a deficiency in the representation
but also prejudice. Strickland v. Washington, 466 U.S. 668, 692 (1984).
For prejudice, the petitioner “must show that there is a reasonable
probability that, but for counsel’s unprofessional errors, the result of the
proceeding would have been different.” Id. at 694.
B. Failure to Seek a Pretrial Hearing on Intellectual Disability
as a Bar to Execution
Mr. Harris argues that his counsel was ineffective for failing to seek
a pretrial hearing on an intellectual disability that would render him
ineligible for the death penalty. This argument is based on Atkins v.
Virginia, 536 U.S. 304 (2002), where the Supreme Court concluded that the
7
execution of intellectually disabled persons violates the Eighth
Amendment’s prohibition on cruel-and-unusual punishment. 536 U.S. at
317, 321. 4
Despite this conclusion, the Supreme Court allowed states to
establish their own standards for an intellectual disability. Id. at 317 n.22.
We thus focus on the content of Oklahoma law (when Mr. Harris’s retrial
took place). At that time, Oklahoma law allowed consideration of an
intellectual disability only if the defendant had at least one IQ score under
70. See Murphy v. State, 54 P.3d 556, 567–68 (Okla. Crim. App. 2002),
overruled in part on other grounds by Blonner v. State, 127 P.3d 1135,
1139 (Okla. Crim. App. 2006). Upon such a showing, the defendant could
then establish an intellectual disability by proving intellectual and adaptive
deficits and manifestation before age eighteen. Id.; see p. 31, below.
Mr. Harris argues that his attorney was ineffective by failing to ask
for a pretrial hearing on intellectual disability. To address this argument,
we consider and apply the standard of review.
4
In the first direct appeal, Mr. Harris’s appellate counsel invoked
Atkins, urging the OCCA to remand for the state trial court to determine
the existence of an intellectual disability. But the OCCA vacated the
sentence without reaching this issue. Harris v. State, 84 P.3d 731, 757
(Okla. Crim. App. 2004).
8
1. The Standard of Review
In denying relief on this claim, the OCCA explained that “[Mr.]
Harris must [1] show that counsel’s performance was so deficient that he
did not have counsel as guaranteed by the Sixth Amendment, and that [2]
the deficient performance created errors so serious as to deprive him of a
fair trial with reliable results.” Harris v. State, 164 P.3d 1103, 1114 (Okla.
Crim. App. 2007). The OCCA rejected this claim on the ground that Mr.
Harris could not establish prejudice. See id. at 1115–16 (concluding that
“Harris cannot show he was prejudiced by counsel’s failure” because “[w]e
cannot conclude there was a reasonable probability that, but for counsel’s
omission, the results of this resentencing proceeding would have been
different”).
The State nevertheless argues that the OCCA implicitly decided the
deficiency prong on the merits. The State’s argument conflates two of the
OCCA’s determinations: One involves Mr. Harris’s claim that his counsel
failed to seek a pretrial hearing on the existence of an intellectual
disability; the other determination involves Mr. Harris’s claim that his
counsel failed to adequately present mitigating evidence at the trial. See
Harris v. State, 164 P.3d 1103, 1118 (Okla. Crim. App. 2007). For the
second claim (failure to adequately present mitigating evidence at the
trial), the OCCA addressed the merits of the deficiency prong. But the
OCCA did not address the deficiency prong on the first claim (failure to
9
seek a pretrial hearing on intellectual disability). For this claim, the OCCA
expressly rested on the prejudice prong without any mention of the
deficiency prong. Harris v. State, 164 P.3d 1103, 1115–16 (Okla. Crim.
App. 2007).
Because the OCCA did not adjudicate the merits of the deficiency
prong on this claim, we engage in de novo review of this part of the
district court’s ruling. See Rompilla v. Beard, 545 U.S. 374, 390 (2005)
(reviewing de novo the prejudice prong of an ineffective-assistance claim
because the state court had not reached this prong); Smith v. Sharp, 935
F.3d 1064, 1072 (10th Cir. 2019) (“[I]n cases in which a state court
addresses only one prong of a multi-prong analysis, the Supreme Court
requires that federal habeas courts address the other prongs de novo.”).
But the OCCA did reach the merits of the prejudice prong, rejecting
Mr. Harris’s arguments. Still, Mr. Harris argues that we should engage in
de novo review on this prong because the OCCA did not
sufficiently consider Dr. Callahan’s report or
permit an evidentiary hearing.
Mr. Harris did not raise his first argument (insufficient consideration
of the evidence by the OCCA) in district court. Even in habeas cases
involving the death penalty, we consider arguments forfeited or waived
when they are raised for the first time on appeal. See Hancock v. Trammell,
10
798 F.3d 1002, 1011 (10th Cir. 2015) (forfeited); Owens v. Trammell, 792
F.3d 1234, 1246 (10th Cir. 2015) (waived). 5
Mr. Harris’s second argument (the OCCA’s denial of an evidentiary
hearing) is based on Wilson v. Workman, 577 F.3d 1284 (10th Cir. 2009)
(en banc), where we considered the OCCA’s denial of an evidentiary
hearing and rejection of an ineffective-assistance claim without
considering material non-record evidence. In these circumstances, we
concluded that the denial did not constitute an adjudication on the merits
under § 2254(d). Wilson, 577 F.3d at 1300.
After we issued this opinion, however, the OCCA clarified its
procedures for deciding these claims. Simpson v. State, 230 P.3d 888
(Okla. Crim. App. 2010). Given this clarification, we concluded in Lott v.
Trammell that
Wilson no longer applies and
any denial of a request for an evidentiary hearing on an
ineffective-assistance claim constitutes an adjudication on the
merits.
5
Our precedents are inconsistent in discussing preservation in cases
involving 28 U.S.C. § 2254. We sometimes treat unpreserved issues as
waived, sometimes as forfeited. See Harmon v. Sharp, 936 F.3d 1044,
1085–91 (10th Cir. 2019) (Holmes, J., concurring) (discussing this
inconsistency in our case law). The difference here is academic. If the
issue involves forfeiture rather than waiver, we could consider the issue
under the plain-error standard. United States v. Carrasco-Salazar, 494 F.3d
1270, 1272 (10th Cir. 2007). But Mr. Harris has not argued plain error, so
we would not entertain the issue even if it had been forfeited rather than
waived. See Hancock, 798 F.3d at 1011.
11
705 F.3d 1167, 1213 (10th Cir. 2013). Mr. Harris’s argument is thus
foreclosed by Lott.
Mr. Harris contends that (1) the panel in Lott could not overrule the
en banc opinion in Wilson and (2) the OCCA’s clarification of the standard
came after the OCCA had rejected Mr. Harris’s argument. We reject both
contentions.
It is true that a panel typically cannot overrule an earlier precedent.
United States v. White, 782 F.3d 1118, 1123 n.2 (10th Cir. 2015). But a
panel is not bound by precedents that have been superseded by a change in
state law. Wankier v. Crown Equip. Corp., 353 F.3d 862, 867 (10th Cir.
2003). Our interpretation of state law changed when the OCCA clarified
the standard for adjudicating a request for an evidentiary hearing. Lott, 705
F.3d at 1213.
As Mr. Harris points out, the OCCA had rejected his argument before
the OCCA clarified the state-law standard. But the same was true in Lott,
and we relied there on the OCCA’s clarification in deciding that the denial
of an evidentiary hearing constituted an adjudication on the merits. Id.
This approach makes sense because the OCCA was clarifying what its rules
had already been and didn’t suddenly start adjudicating the merits when
denying evidentiary hearings. Wilson v. Trammell, 706 F.3d 1286, 1311
(10th Cir. 2013) (Gorsuch, J., concurring). Before Lott, we had simply
12
misunderstood Oklahoma law. See id. (“[T]he OCCA has explained that
Wilson was mistaken in its understanding of Oklahoma law.”). Under Lott,
we thus consider the OCCA’s denial of an evidentiary hearing on an
ineffective-assistance claim as an adjudication on the merits.
We thus engage in de novo review of the OCCA’s ruling on the
deficiency prong, but we apply § 2254(d)’s deferential standard of review
on the prejudice prong.
2. Deficiency Prong
Applying de novo review, we conclude that Mr. Harris’s attorney was
deficient in failing to request a pretrial hearing to assess an intellectual
disability.
The State argues that defense counsel strategically decided to forgo a
pretrial hearing after a thorough investigation. Strategic decisions draw
considerable deference when the attorney has thoroughly investigated the
law, the facts, and the plausible alternatives. Strickland v. Washington, 466
U.S. 668, 690 (1984). But merely calling something a strategy does not
prevent meaningful scrutiny. We must still determine (1) whether an
attorney has chosen to forgo a course of action and (2) whether that choice
was reasonable under the circumstances. Brecheen v. Reynolds, 41 F.3d
1343, 1369 (10th Cir. 1994).
In assessing the reasonableness of an attorney’s investigation, we
engage in close scrutiny during the penalty phase of capital cases.
13
Littlejohn v. Trammell, 704 F.3d 817, 859 (10th Cir. 2013). In these cases,
“we refer to the ABA Guidelines for the Appointment and Performance of
Counsel in Death Penalty Cases.” Id. These guidelines require that
“[c]ounsel at every stage of the case should take advantage of all
appropriate opportunities to argue why death is not suitable punishment for
their particular client.” ABA Guidelines § 10.11(L). One appropriate
opportunity involved a pretrial hearing on the existence of an intellectual
disability. 6 State ex rel. Lane v. Bass, 87 P.3d 629, 633 (Okla. Crim. App.
2004), overruled in part on other grounds by Blonner v. State, 127 P.3d
1135, 1139 (Okla. Crim. App. 2006). Had Mr. Harris been found
intellectually disabled, he would have been ineligible for the death penalty.
Id. at 632.
When the 2005 retrial took place, Oklahoma law permitted pretrial
evidentiary hearings before a judge on the existence of an intellectual
disability. See State ex rel. Lane v. Bass, 87 P.3d 629, 633–35 (Okla. Crim.
App. 2004), overruled in part on other grounds by Blonner v. State, 127
P.3d 1135, 1139 (Okla. Crim. App. 2006). If the defendant preferred a
jury, he or she could also opt for a jury finding on the existence of an
6
Alternatively, Mr. Harris could have asked the trial jury to determine
the existence and impact of an intellectual disability. Lane, 87 P.3d at 632.
But Mr. Harris argues only that his attorney should have requested a
pretrial hearing.
14
intellectual disability. If the jury found no intellectual disability, the
defendant could ask the judge to revisit the issue after the trial. Id. at 635.
So if the judge or jury found no intellectual disability, the defense
would have lost nothing. But if either the judge or jury found an
intellectual disability, the death penalty would have vanished as a
possibility. Defense counsel thus had a risk-free opportunity to avoid the
death penalty. Frazier v. Jenkins, 770 F.3d 485, 501 (6th Cir. 2014) 7; see
Clinkscale v. Carter, 375 F.3d 430, 443 (6th Cir. 2004) (holding that
defense counsel was deficient by failing to file a timely notice of an alibi
defense when counsel had “everything to gain” and “nothing to lose”); see
also Browning v. Baker, 875 F.3d 444, 473 (9th Cir. 2017) (“[T]he
obligation to investigate, recognized by Strickland, exists when there is no
reason to believe doing so would be fruitless or harmful.”). 8
7
The Frazier court explained: “[W]e fail to see the downside in
having a non-frivolous Atkins hearing, and it is difficult to ascertain a
strategic reason for withdrawing the motion [for an Atkins hearing] in this
case.” 770 F.3d at 501.
8
At oral argument, the State also suggests that Mr. Harris might have
wanted to avoid the delay from a pretrial hearing on intellectual disability.
But the State had never before argued in state or federal court that Mr.
Harris wanted to expedite his capital proceedings. See United States v.
Gaines, 918 F.3d 793, 800–801 (10th Cir. 2019) (“We typically decline to
consider an appellee’s contentions raised for the first time in oral
argument.”).
15
Though no downside existed, 9 a pretrial hearing had considerable
upside. The evidence of an intellectual disability was ready-made. For
example, Mr. Harris had IQ scores under the 70-point threshold necessary
for a determination of intellectual disability under Oklahoma law. Murphy
v. State, 54 P.3d 556, 567–68 (Okla. Crim. App. 2002), overruled in part
on other grounds by Blonner v. State, 127 P.3d 1135, 1139 (Okla. Crim.
App. 2006). One expert witness, Dr. Martin Krimsky, had already
diagnosed Mr. Harris with a mild intellectual disability. And other
evidence of Mr. Harris’s difficulties in intellectual and adaptive
functioning had already been introduced at a competency hearing and the
2001 trial.
The State contends that defense counsel did not request a pretrial
hearing because he believed that Mr. Harris was not intellectually
disabled. 10 For this contention, the State points to the voir dire, where
9
We do not suggest that counsel should always argue points lacking
any downside. See Knowles v. Mirzayance, 556 U.S. 111, 121–22 (2009)
(stating that counsel may not be ineffective by declining to assert a defense
even when there is nothing to lose).
10
In oral argument, the State also argues for the first time that a
pretrial hearing on intellectual disability might have generated new
evidence for the State to support an aggravating circumstance. This
argument was omitted in the briefs. See note 8, above. But even if we were
to consider this argument, the State does not explain what new evidence
would have been elicited at the pretrial hearing that had not already been
fully aired in the 2001 proceedings. Those proceedings included a
competency hearing and trial, and both included considerable evidence of
Mr. Harris’s mental state. In fact, the State ultimately conceded that any
16
defense counsel conceded that Mr. Harris was not intellectually disabled.
We do not know why defense counsel made this concession, 11 and there is
nothing to suggest that he had investigated the possibility of an intellectual
disability. Before this concession, Dr. Krimsky had already testified that
Mr. Harris was intellectually disabled. Even if defense counsel had
disagreed with Dr. Krimsky’s assessment, the ABA guideline required him
to take advantage of every opportunity to argue against a death sentence.
One such opportunity existed for a pretrial hearing on an intellectual
disability, and the failure to request this hearing fell outside the acceptable
range of reasonable performance. See Williamson v. Ward, 110 F.3d 1508,
1517–18 & n.12 (10th Cir. 1997) (concluding that the petitioner’s counsel
was ineffective in failing to seek a competency hearing given the existing
evidence of incompetency and the lack of any strategic advantage).
* * *
Defense counsel had nothing to lose by requesting a pretrial hearing
on an intellectual disability. Prevailing would have eliminated the
possibility of the death penalty, and losing would have left Mr. Harris
resulting evidence in aggravation had already been created in the 2001
proceedings. We thus reject the State’s eventual argument that the pretrial
hearing might have generated additional evidence of an aggravating
circumstance.
11
When defense counsel made the concession, he was supposed to be
asking questions to the venirepersons.
17
precisely where he would be anyway, free to urge acquittal and a life
sentence upon a conviction. Given the evidence already developed in the
2001 proceedings, any reasonable defense attorney would have sought a
pretrial hearing on the existence of an intellectual disability. By failing to
seek a pretrial hearing, Mr. Harris’s attorney bypassed a risk-free
opportunity to avoid the death penalty. Bypassing this opportunity
constituted a deficiency in the representation.
3. Prejudice Prong
Because the OCCA adjudicated the prejudice prong on the merits, the
federal district court could have reached the merits of the prejudice issue
only if Mr. Harris had cleared the hurdle under 28 U.S.C. § 2254(d). See
pp. 3–5, above. Section 2254(d) prevents consideration of the merits unless
the OCCA’s decision on prejudice was (1) contrary to, or an unreasonable
application of, clearly established federal law or (2) based on an
unreasonable determination of fact in light of the evidence presented in
state court. 28 U.S.C. § 2254(d)(1)–(2).
In our view, the OCCA’s decision on prejudice was based on an
unreasonable factual determination, so we consider the merits. 12
12
Given this conclusion, we need not decide whether the OCCA’s
decision on prejudice was contrary to, or an unreasonable application of,
clearly established federal law.
18
(a) Unreasonable Determination of Fact
Mr. Harris contends that the OCCA’s decision was based on an
unreasonable factual determination under 28 U.S.C. § 2254(d)(2). He
points to this passage in the OCCA’s decision: “All Harris’s experts,
including the ones who testified at his [2001] trial and competency
hearing, considered these scores along with Harris’s other characteristics
and concluded he was not mentally retarded.” Harris v. State, 164 P.3d
1103, 1115 (Okla. Crim. App. 2007). Mr. Harris contends that this passage
reflects an unreasonable determination of fact because Dr. Krimsky had
assessed an intellectual disability. 13
The State argues that Mr. Harris failed to preserve this contention in
district court by limiting his argument to Dr. Callahan’s affidavit. We
disagree.
To preserve the issue in district court, Mr. Harris needed only to
alert the court to the issue and seek a ruling. See Ecclesiastes 9:10-11-12,
Inc. v. LMC Holding Co., 497 F.3d 1135, 1141 (10th Cir. 2007) (“An issue
is preserved for appeal if a party alerts the district court to the issue and
seeks a ruling.”); United States v. Harrison, 743 F.3d 760, 763 (10th Cir.
2014) (stating that the test for specificity of an objection in district court
13
Dr. Krimsky actually used the term “mentally retarded.” But in
analyzing Mr. Harris’s claim, we use the term “intellectually disabled.”
See note 1, above.
19
“is whether the district court was adequately alerted to the issue”). We thus
consider whether Mr. Harris’s argument in district court encompassed Dr.
Krimsky’s opinion. The State answers “no;” we answer “yes.”
In district court, Mr. Harris treated Dr. Callahan’s opinion as
significant new evidence of intellectual disability. But Mr. Harris did not
confine his argument to Dr. Callahan’s opinion. Mr. Harris’s argument on
prejudice spanned roughly 32 pages. Within this discussion lay Mr.
Harris’s challenge to the OCCA’s characterization of the expert opinions.
Mr. Harris prefaces this discussion by explaining why the OCCA’s
decision was unreasonable under § 2254(d)(2). See Habeas Pet. at 107
(“Below is a discussion of the three (3) criteria, the impact of Dr.
Callahan’s report and argument why the OCCA decision was unreasonable
under both prongs of §2254 (d).”). In the ensuing section, Mr. Harris
extensively discusses all of the prior expert opinions on the existence of an
intellectual disability.
For example, in discussing the criterion of significant sub-average
intellectual functioning, Mr. Harris discusses Dr. Callahan’s references to
IQ tests administered by herself, Dr. Martin Krimsky, and Dr. Nelda
Ferguson. Mr. Harris notes that the IQ tests by Dr. Ferguson and Dr.
Krimsky would have met the state-law criterion for IQ test results below
70. And Mr. Harris underscores Dr. Krimsky’s test results and assessment
of mild intellectual disability:
20
Dr. Krimsky concluded the IQ scores indicated that Mr. Harris
was mildly mentally retarded. He did not believe Mr. Harris was
malingering or “trying to fool the test.” He again confirmed
Jimmy Dean Harris “an individual with mental retardation.”
Id. at 110 (citations omitted).
Mr. Harris also discusses the expert opinions by Dr. John Smith, Dr.
Wanda Draper, and Dr. Ray Hand. In this discussion, Mr. Harris points out
that Dr. Smith confirmed Dr. Krimsky’s testing as an indication of
intellectual disability. Id. at 112.
Despite this broad record-based attack on the OCCA’s factual
determination, the State points to two pages in which Mr. Harris discusses
his reliance on Dr. Callahan’s opinion. The State’s reliance on these two
pages disregards the other 30 pages in Mr. Harris’s argument as well as the
nature of Dr. Callahan’s report. In this report, Dr. Callahan relied not only
on her own examination and testing but also on the prior testing and
diagnoses. For example, Dr. Callahan noted that Dr. Krimsky, Dr.
Ferguson, and Dr. Smith had separately diagnosed Mr. Harris as having a
mild intellectual disability.
The State also argues that Mr. Harris was relying solely on Dr.
Callahan’s opinion. We disagree. Mr. Harris addressed all of the expert
witnesses, including both Dr. Krimsky and Dr. Callahan. On appeal, Mr.
Harris narrows his focus to Dr. Krimsky. This narrower argument is
subsumed by the broader argument that Mr. Harris had presented in district
21
court. The district court was thus alerted to Mr. Harris’s appellate
argument, which sufficed for preservation. See Joseph A. ex rel. Wolfe v.
N.M. Dep’t of Hum. Servs., 28 F.3d 1056, 1060 (10th Cir. 1994)
(concluding that the appellants had preserved their appellate argument
because it had been subsumed by the argument presented in district court);
accord PCTV Gold, Inc. v. Speednet, LLC, 508 F.3d 1137, 1144 n.5 (8th
Cir. 2007) (concluding that an appellate argument was preserved because it
had been encompassed in a more general argument presented in district
court). Because Mr. Harris preserved the issue, we consider the merits of
his challenge to the reasonableness of the OCCA’s factual determination.
We conclude that the OCCA was clearly mistaken as to Dr. Krimsky.
The OCCA concluded that all of the defense experts had opined that Mr.
Harris was not intellectually disabled. Harris v. State, 164 P.3d 1103, 1115
(Okla. Crim. App. 2007). But Dr. Krimsky had opined that Mr. Harris was
intellectually disabled.
In our appeal, the State appears to acknowledge expert testimony that
Mr. Harris is intellectually disabled: “The only experts who have opined
that Petitioner is mentally retarded have relied upon unreliable test results
that contradict the experts’ experiences with him.” Appellee’s Resp. Br. at
32–33. In oral argument, the State elaborates on this argument, insisting
that the OCCA could reasonably reject Dr. Krimsky’s test results because
Mr. Harris was psychotic at the time of testing. But this was not the
22
OCCA’s rationale. The OCCA reasoned that all defense experts had opined
that Mr. Harris was not intellectually disabled, and this was simply not
true of Dr. Krimsky. Harris, 164 P.3d at 1115.
The State also denies that the OCCA misunderstood Dr. Krimsky’s
opinion. The State points to a footnote where the OCCA
noted that one expert had believed that he “had” to say that Mr.
Harris’s test scores indicated an intellectual disability but
added that it “was not his conclusion” after examining Mr.
Harris.
Id. at 1115 n.55.
The State’s argument misstates the testimony. Dr. Krimsky testified
that he had administered two IQ tests: (1) the Slossen Intelligence Test
Revised (“SIT”) and (2) the Wechsler Adult Intelligence Scale, Revised
(“WAIS-R”). Mr. Harris scored a 66 on the SIT and a 68 on the WAIS-R,
and Dr. Krimsky regarded these scores as proof of mild intellectual
disability.
He explained that “[t]here was an ambiguity comparing the result of
the first test [the SIT] . . . and [Mr. Harris’s] occupation of having been
involved in repair of auto transmissions.” 2001 Comp. Hearing, vol. 1, at
63. But Dr. Krimsky noted that the second test [the WAIS-R] was “much
more comprehensive” with “a high validity in relation to occupational and
socioeconomic status.” Id. at 64. Dr. Krimsky ultimately considered both
sets of results to be consistent and accurate.
23
Dr. Krimsky also testified that Mr. Harris’s mechanical skills could
have been acquired by someone who was mildly intellectually disabled,
pointing out that Mr. Harris had spent “a long period of time . . . observing
his father and other people fix transmissions.” Id. at 65. Given this lengthy
period of observation, Dr. Krimsky opined that Mr. Harris’s low IQ was
consistent with his skill in fixing transmissions.
Dr. Krimsky thus testified that Mr. Harris’s skills did not undermine
the assessment of mild intellectual disability. In fact, Dr. Krimsky
corrected an attorney who had referred to Mr. Harris as “borderline,” with
Dr. Krimsky repeating his characterization of Mr. Harris as having “mild
mental retardation.” 14 Id. The OCCA thus made an unreasonable factual
finding that all of Mr. Harris’s experts had opined that he was not
14
At one point, Dr. Krimsky was asked, “[W]hat conclusions did you
come to regarding [Mr. Harris’s] mental state as far as his IQ and the
mental retardation?” Id. He answered that the “mental retardation” was
“incidental.” Id. In Dr. Krimsky’s view, Mr. Harris was “in a psychotic
status and in need of mental health treatment, psychiatric treatment.” Id. at
65–66. Dr. Krimsky used the term “mental state” to refer to Mr. Harris’s
competency and his ability to retain consistent contact with his “outer
situation.” Id. at 66–67. With respect to this mental state, Dr. Krimsky
concluded that Mr. Harris was delusional and not competent, adding that
Mr. Harris’s competency could probably be restored within a reasonable
period of time. But Dr. Krimsky did not testify that the delusions had
affected the IQ scores or that Mr. Harris was trying to manipulate the
results. Indeed, Dr. Krimsky’s assessment of Mr. Harris’s intellectual
disability remained consistent throughout the competency hearing. In Dr.
Krimsky’s unchanging view, Mr. Harris had mild intellectual disability.
24
intellectually disabled. Dr. Krimsky was one of Mr. Harris’s experts, and
he specifically opined that Mr. Harris was intellectually disabled.
The State also argues that even if the OCCA’s factual determination
had been unreasonable, this factual determination had not formed the basis
for the OCCA’s decision. As the State points out, it is not enough for Mr.
Harris to show an unreasonable factual determination; the state court’s
decision must have also been “based on” the unreasonable factual
determination. Byrd v. Workman, 645 F.3d 1159, 1172 (10th Cir. 2011).
In our view, however, the OCCA did indeed base its decision on the
unreasonable factual determination. The OCCA explained that it had found
no prejudice:
Nothing in this record shows that, had counsel made [a request
for a pretrial hearing], evidence would have shown by a
preponderance of the evidence that Harris was mentally retarded.
There is a great deal of evidence in the record to show otherwise,
including the opinion of several experts who testified that Harris
was not mentally retarded. We cannot conclude that there was a
reasonable probability that, but for counsel’s omissions, the
results of this resentencing proceeding would have been
different.
Harris v. State, 164 P.3d 1103, 1116 (Okla. Crim. App. 2007) (emphasis
added). 15 By highlighting the expert opinions rejecting an intellectual
15
In assessing the evidence, the OCCA disregarded the fact that the
controlling Oklahoma definition of intellectual disability was set forth in a
case decided after the competency hearing and the first trial. Murphy v.
State, 54 P.3d 556, 567¬68 (Okla. Crim. App. 2002), overruled in part on
other grounds by Bloomer v. State, 127 P.3d 1135, 1139 (Okla. Crim. App.
2006). Accordingly, none of the 2001 testimony applied the controlling
25
disability, the OCCA suggested that this was the critical evidence on
prejudice. The OCCU thus based its decision on its perception of the
various expert opinions, including its mistaken perception of Dr.
Krimsky’s opinion.
(b) The Need for an Evidentiary Hearing
We thus must tackle the prejudice prong in the first instance. Magnan
v. Trammell, 719 F.3d 1159, 1175 (10th Cir. 2013). To do so, we must
consider the evidence of intellectual disability.
standard for an intellectual disability. We have no way of knowing what
the expert witnesses would have said if they had applied the standard for
an intellectual disability that governed at the time of the 2005 retrial. For
example, Dr. Ray Hand testified at the first trial that Mr. Harris had
exhibited “borderline intellectual functioning” but was not “mentally
retarded.” 2001 Tr., v. 15, at 133–34. But Dr. Hand based that conclusion
in part on his view about which IQ scores were “more realistic and more
representative of [Mr. Harris’s] actual abilities.” Id. at 131. In contrast,
the controlling standard does not require the parties or the court to identify
the more realistic or representative score. The question is instead whether
the defendant has “an intelligence quotient of seventy or below, as
reflected by at least one scientifically recognized, scientifically approved,
and contemporary intelligence quotient test.” Murphy, 54 P.3d at 568. Dr.
Hand did not apply this test.
Dr. Hand also testified about various deficits in Mr. Harris’s
adaptive functioning. But Dr. Hand was not asked whether Mr. Harris had
“ significant limitations in adaptive functioning in at least two of the
following skill areas: communication; self-care; social/interpersonal skills;
home living; self-direction; academics; health and safety; use of
community resources; and work.” Murphy, 54 P.3d at 568 .
26
Mr. Harris contends that a pretrial hearing could have led to a
finding of intellectual disability, pointing to his history of IQ testing, Dr.
Callahan’s report, expert testimony, and evidence of difficulties in
adaptive functioning. In response, the State focuses on Mr. Harris’s older
IQ tests, the testimony of other experts, and Mr. Harris’s employment
history.
The issue of prejudice turns on whether a reasonable factfinder could
find an intellectual disability. With this issue hotly disputed and the lack
of a factual finding, the district court could not grant habeas relief. See
Littlejohn v. Trammell, 704 F.3d 817, 856 (10th Cir. 2013) (stating that
even though counsel’s conduct may have been prejudicial, the court could
not grant habeas relief “[a]t this juncture” because the persuasiveness of
particular expert testimony was disputed and the claim was “highly fact-
bound”).
Nor could the district court deny habeas relief, for no factfinder has
considered Mr. Harris’s evidence of intellectual disability based on the
Oklahoma test that applied during Mr. Harris’s retrial. Without a factual
finding based on the applicable test, a court could not properly assess the
extent of the prejudice.
To decide the issue of prejudice, the district court needed to assess
the likelihood that defense counsel could have proven the existence of an
intellectual disability. Like us, the district court had only a cold record
27
containing conflicting evidence on Mr. Harris’s intellectual status. Dr.
Krimsky assessed an intellectual disability; Dr. Callahan assessed
borderline intellectual functioning; and Dr. Draper considered Mr. Harris
to be intellectually impaired but not intellectually disabled. 16
No court has had the opportunity to hear these experts testify and
apply the Oklahoma test on intellectual disability. If these experts had
testified in a pretrial hearing focused on that test, which experts would
have swayed the factfinder? To provide at least a meaningful prediction, a
court must at least hear the conflicting evidence, apply Oklahoma’s test for
an intellectual disability, and determine which expert witnesses to believe.
See Smith v. Sharp, 935 F.3d 1064, 1077 (10th Cir. 2019) (stating that
“Atkins clearly establishes that intellectual disability must be assessed, at
least in part, under the existing clinical definitions applied through expert
testimony” and recognizing “the centrality of expert testimony to our
review of Atkins verdicts”). No court has engaged in this scrutiny, so any
court would need an evidentiary hearing to predict the outcome of a
pretrial hearing on an intellectual disability.
16
Dr. Hand and Dr. Smith supplied other assessments. Dr. Hand did not
believe that Mr. Harris was mentally retarded (under his definition of
mental retardation) but thought that he had “mixed specific learning
disabilities” and was likely “slow” or had “borderline intellectual
functioning.” 2001 Tr., v. 15, at 133–34. And Dr. Smith believed that Mr.
Harris had “normal intelligence.” Comp. Hearing, v. 1, at 215.
28
We addressed a similar situation in Littlejohn v. Trammell, 704 F.3d
817 (10th Cir. 2013). There we concluded that the availability of habeas
relief turned on a disputed factual issue that prevented a meaningful
decision based on the cold record alone. Littlejohn, 704 F.3d at 856. We
directed the district court to conduct an evidentiary hearing on the issue of
prejudice. Id. Here we have the same need for an evidentiary hearing.
An evidentiary hearing is ordinarily unavailable when the petitioner
failed to diligently develop the factual bases of the claim in state court.
Williams v. Taylor, 529 U.S. 420, 432 (2000). 17 Here, however, Mr. Harris
diligently tried to develop the factual foundations of his claim when he
was in state court. For example, he argued that his trial counsel had failed
to seek a pretrial hearing on intellectual disability. With this argument, Mr.
Harris requested an evidentiary hearing and supported the request with Dr.
Callahan’s affidavit. The OCCA denied this request.
17
Exceptions exist when the habeas claim is based on
a new constitutional rule that the Supreme Court has made
retroactive on collateral review or
a factual predicate not reasonably discoverable earlier through
reasonable diligence, along with clear and convincing evidence
showing that no reasonable factfinder would have found guilt
without the constitutional error.
28 U.S.C. § 2254(e)(2)(A)(i)–(ii).
29
Mr. Harris did all that he could to develop the factual foundation for
a showing of prejudice. By denying the opportunity for an evidentiary
hearing, the OCCA left us with only a cold record and no factual findings
for the innately fact-intensive issue of prejudice.
Because Mr. Harris was diligent, we consider whether Mr. Harris’s
proof of allegations would entitle him to habeas relief. See Hammon v.
Ward, 466 F.3d 919, 927 (10th Cir. 2006). That inquiry turns on the issue
of prejudice. Defense counsel’s deficient performance would be prejudicial
if a pretrial hearing would create a reasonable probability of a lesser
sentence. See Strickland v. Washington, 466 U.S. 668, 694 (1984).
Mr. Harris argues that if his trial attorney had requested a pretrial
hearing, the trial court would have granted the request and found Mr.
Harris intellectually disabled, rendering him ineligible for execution. We
thus gauge the likelihood that the state court would have found an
intellectual disability.
As noted, the Supreme Court has prohibited the execution of
intellectually disabled individuals, but allowed the states to define the term
“intellectual disability.” Atkins v. Virginia, 536 U.S. 304, 317 (2002).
When Mr. Harris appealed his conviction, Oklahoma law required a
defendant to show at least one IQ score under 70. Murphy v. State, 54 P.3d
556, 567–68 (Okla. Crim. App. 2002), overruled in part on other grounds
by Blonner v. State, 127 P.3d 1135, 1139 (Okla. Crim. App. 2006). If the
30
defendant produced at least one score under 70, he or she would need to
satisfy three elements:
1. The person “functions at a significantly sub-average
intellectual level that substantially limits his or her ability to
understand and process information, to communicate, to learn
from experience or mistakes, to engage in logical reasoning, to
control impulses, and to understand the reactions of others.”
2. The disability “manifested itself before the age of eighteen.”
3. The disability “is accompanied by significant limitations in
adaptive functioning in at least two of the following skill areas:
communication; self-care; social/interpersonal skills; home
living; self-direction; academics; health and safety; use of
community resources; and work.”
Id.; see p. 8, above.
Mr. Harris’s counsel could have satisfied the threshold requirement
for at least one IQ score below 70. And the State does not challenge the
second element (manifestation before the age of eighteen). The dispute
exists on the first and third elements, which address Mr. Harris’s
intellectual and adaptive deficits.
Mr. Harris’s evidence on intellectual deficits involves three
categories:
1. his history of IQ testing,
2. the testimony of an expert witness, and
3. the affidavit of an expert witness.
First, Mr. Harris’s IQ testing began in his childhood. Two childhood
IQ tests yielded scores of 87 and 83. After the murder, new IQ tests
31
yielded scores of 63, 66, 68, and 75. And after Mr. Harris’s retrial, Dr.
Jennifer Callahan tested Harris’s IQ and obtained scores ranging from 67–
75 and 72–77.
Date Type of Test By Score
1964 Stanford- Dr. Teresa Costiloe at 87
Binet Revised University of Oklahoma
Hospital
1964 WISC Dr. Teresa Costiloe at 83
University of Oklahoma
Hospital
The Murder
Oct. WAIS-III Dr. Nelda Ferguson 63
20,
2000
March SIT-R Dr. Martin Krimsky 66
8,
2001
March WAIS-R Dr. Martin Krimsky 68
21,
2001
July WAIS-III Dr. Elizabeth Grundy at 75
20, Eastern State Hospital
2001
Sentencing, Resentencing, and Direct Appeal
March WASI-I Dr. Jennifer Callahan 67–75
13,
2006
March Woodcock- Dr. Jennifer Callahan 72–77
13, Johnson III
2006
Second, Mr. Harris points to Dr. Krimsky’s testimony about his two
IQ tests. Dr. Krimsky testified in the 2001 competency hearing, explaining
that his testing showed “mild mental retardation.” See 2001 Comp.
Hearing, v. 1, at 58. When asked whether Mr. Harris’s occupation was
32
consistent with borderline intellectual disability, Dr. Krimsky corrected the
attorney, pointing out that Mr. Harris was “not borderline” and reiterated
that he had “mild mental retardation.” Id. at 65.
Third, Mr. Harris points to an affidavit and report by Dr. Callahan,
who concluded that Mr. Harris’s IQ fell in the “impaired to borderline
impaired range.” R. at 287. On one test, Mr. Harris’s scaled score was 67–
75; on another test, the scaled scored was 72–77, which Dr. Callahan said
would approximate the mental status of a child only 6 years and 10 months
old. Dr. Callahan explained the disparity in Mr. Harris’s IQ scores,
concluding that “greater consistency” existed in the scores than “one may
appreciate initially” because IQ is ideally viewed as a range and IQ scores
change over time based on a phenomenon known as the “Flynn effect.” Id.
at 288.
The Flynn effect is designed to account for two facts:
1. IQ tests measure intelligence relative to the contemporaneous
general population, not as an absolute number.
2. IQ scores tend to increase over time.
Given these two facts, an older IQ test would typically yield a higher
figure than a more recent test for the same individual. For example, Mr.
Harris took one of the IQ tests in 1964. By the time of Mr. Harris’s test,
the grading scale was roughly fifteen years old. So Dr. Callahan lowered
Mr. Harris’s score from 83 +/- 5 to 75.5 +/- 5.
33
Dr. Callahan concluded that her findings indicated “borderline
intellectual functioning,” but she acknowledged that Mr. Harris’s cognitive
abilities were “not uniformly at this level.” Id. at 289. 18
Mr. Harris also presented six forms of evidence involving adaptive
deficits:
1. Dr. Callahan’s testing showed adaptive strengths, including Mr.
Harris’s “visual-spatial thinking abilities,” which explained
how he could work. But his “relative weakness[es]” included
the inability to quickly process information, difficulty in
learning and recalling new information, and impairment in his
ability to “plan and organize.” Id. at 289–90.
2. Mr. Harris had a history of poor academic performance. Even
with tutors, he dropped out of high school and experienced
problems in recognizing words, spelling, and doing
mathematics. These problems led Dr. Callahan to regard Mr.
Harris as functionally illiterate, with abilities approximating
those of a first or second grader.
3. Though Mr. Harris worked as a mechanic, he was “slow” and
his wife needed to read the technical manuals and call hotlines
for help. 2005 Tr., v. 5, at 55–56, 58, 157.
4. A former employer testified about difficulty in communicating
with Mr. Harris, stating that “[h]e would start one sentence and
end it with a different sentence.” 2001 Tr., v. 12, at 28–29.
5. Mr. Harris engaged in very risky behavior as a child and teen,
leading to injuries.
6. Mr. Harris had a lifelong addiction to alcohol and narcotics,
showing difficulties in self-care (a feature of adaptive
functioning).
18
Dr. Callahan added that Mr. Harris was not malingering.
34
This combination of evidence could lead to a reasonable finding that Mr.
Harris had satisfied the first and third elements of an intellectual disability
(impairments in intellectual and adaptive functioning).
The State disagrees, relying on Mr. Harris’s childhood IQ tests and
employment history. But the tests and employment history invoked by the
State are controverted by
1. Dr. Callahan’s discussion of the Flynn effect, which would
contextualize the IQ scores stressed by the State,
2. expert testimony that an intellectual disability would not
necessarily prevent work as a mechanic, and
3. OCCA decisions in other cases stating that similar evidence of
adaptive functioning and borderline intellectual functioning did
not preclude relief. 19
Thus, proof of Mr. Harris’s allegations would support the finding of an
intellectual disability. Given the potential for this finding, a habeas court
could view defense counsel’s failure to request a pretrial hearing as
prejudicial.
Ultimately, however, we cannot accurately resolve the dispute over
the first and third elements of an intellectual disability. Mr. Harris and the
19
For example, in Pickens v. State, 126 P.3d 612 (Okla. Crim. App.
2005), the OCCA concluded that a petitioner was intellectually disabled as
a matter of law when his IQ testing indicated borderline intellectual
functioning and showed some ability to function adaptively. 126 P.3d at
618–20.
35
State point to evidentiary disputes on these elements, and these disputes
have not been presented to a factfinder for resolution under Oklahoma’s
test for an intellectual disability. So a decision on the prejudice prong
should await an evidentiary hearing in district court. See p. 29, above
(discussing Littlejohn v. Trammell, 704 F.3d 817, 856–57 (10th Cir.
2013)); accord Sasser v. Hobbs, 735 F.3d 833, 850 (8th Cir. 2013)
(concluding that “misconceptions about the Arkansas legal standard [for
identifying an intellectual disability] led the district court to answer the
wrong factual questions, leaving the pertinent questions unanswered” and
that “[t]he proper course . . . [was] to vacate the district court’s finding
that [the defendant] [was] not mentally retarded and remand so that the
district court [could] answer the critical factual questions in the first
instance according to the correct legal standard”); Allen v. Buss, 558 F.3d
657, 663 (9th Cir. 2009) (observing that “the [state] trial court did not
determine whether [the petitioner] is mentally retarded under Indiana’s test
for mental retardation” and remanding the case to the federal district court
for an evidentiary hearing).
(c) Conclusion
Engaging in de novo review, we conclude that Mr. Harris has
shown a deficiency in defense counsel’s performance and
alleged a theory of prejudice that, if true, could justify habeas
relief.
36
Although factual disputes preclude us from deciding the issue of prejudice,
Mr. Harris is entitled to an evidentiary hearing. We thus remand for an
evidentiary hearing as to prejudice. At this hearing, the parties should be
able to present expert testimony on whether Mr. Harris satisfied
Oklahoma’s test for an intellectual disability. Smith v. Sharp, 935 F.3d
1064, 1077 (10th Cir. 2019) (noting our prior recognition of “the centrality
of expert testimony to our review of Atkins verdicts”).
C. Failure to Adequately Present Mitigation Evidence
The Supreme Court has recognized that attorneys in death-penalty
cases are ineffective if they bypass evidence that might have altered the
jury’s selection of a penalty. Williams v. Taylor, 529 U.S. 362, 398 (2000).
Mr. Harris invokes this case law, arguing that his attorney failed to
adequately present mitigation evidence on intellectual impairments and
mental illness.
Mr. Harris’s arguments encompass evidence that would show not
only an intellectual disability but also lesser intellectual impairments that
the jury could regard as mitigating. Mr. Harris also points to evidence of
other mental illnesses. 20
20
We consider three categories of mitigating evidence. The first is an
“intellectual disability,” meaning evidence that meets the Oklahoma test at
the time of the 2005 retrial. The second is “borderline intellectual
functioning,” which consists of lesser cognitive and adaptive impairments
that might be mitigating. See 2001 Tr., v. 15, at 133–36 (testimony of Dr.
37
1. The Legal Standard and the Standard of Review
For these arguments, we consider whether the OCCA unreasonably
applied Strickland v. Washington, 466 U.S. 668 (1984). To assess
prejudice, we must evaluate the totality of the evidence, including
1. the aggravating circumstances found by the jury,
2. the mitigation evidence,
3. the mitigation evidence that might have been introduced, and
4. “what the prosecution’s response to that evidence would have
been.”
Littlejohn v. Royal, 875 F.3d 548, 553 (10th Cir. 2017), cert. denied, 139
S. Ct. 102 (2018). Applying both prongs of Strickland (deficiency and
prejudice), the OCCA rejected the mitigation-related claims on the
merits. 21 Harris v. State, 164 P.3d 1103, 1118 (Okla. Crim. App. 2007). We
thus apply the standard set out in 28 U.S.C. § 2254(d). See pp. 3–4, above.
2. Intellectual Impairment as a Mitigating Factor
Mr. Harris argues that his attorney performed deficiently by calling
only one expert witness (Dr. Draper) to testify about an intellectual
Hand). The third category consists of other mental illnesses that might be
mitigating.
21
Mr. Harris insists that the OCCA’s denial of an evidentiary hearing
could not have constituted a denial on the merits. But as we explain above,
this argument is based on a misunderstanding of Oklahoma law. See pp.
11–13, above.
38
impairment involving either an intellectual disability or borderline
intellectual functioning. According to Mr. Harris, his attorney should have
presented better mitigation evidence of an intellectual impairment. In our
view, however, the OCCA acted reasonably in rejecting this claim based on
a failure to show either deficient performance or prejudice.
(a) Evidence of Intellectual Impairments
In the 2005 retrial, defense counsel presented testimony by seven of
Mr. Harris’s family, friends, and associates. But Mr. Harris’s attorney
called only one expert witness, Dr. Wanda Draper. Dr. Draper was not an
expert in intellectual impairments; her expertise instead involved
development, an interdisciplinary field involving psychology, sociology,
and other disciplines. She testified mainly about Mr. Harris’s “life path,”
which included his childhood, education, and personal relationships. 2005
Tr., v. 5, at 35.
Some of Dr. Draper’s testimony concerned Mr. Harris’s intellectual
impairments. For example, Dr. Draper testified that Mr. Harris had “[p]oor
[s]chool [p]erformance,” was “[s]low [i]n school,” had an IQ score in the
80s, suffered from “[d]yslexia,” had a “[c]ompulsive personality,” and
experienced a “[p]erception disorder.” Def. Exh. 2. Dr. Draper added that
(1) Mr. Harris’s dyslexia had impeded his ability to read and write and (2)
he had suffered from a “perception disorder,” which led to compulsiveness
and an inability to see things in perspective. 2005 Tr., v. 5, at 43.
39
Dr. Draper explained that Mr. Harris “was not retarded, but he was
slow and he had to do things very slowly and with help.” Id. at 58.
According to Dr. Draper, the need to act slowly rendered him dependent on
Pam. Dr. Draper also explained the unevenness in Mr. Harris’s IQ test
results:
Q: [H]e was given IQ tests, for lack of a better term,
intelligence test, after the fact, after he was arrested.
A: Right.
Q: And there was a scatter in those IQ tests?
A: Yes, they were relatively low, but there was a scatter. And
because he had what we would call high level of spatial
and visual intelligence he was able to do that transmission
work. He had good eye/hand coordination. And he was able
to look at a three-dimensional object and figure how it goes
together in a car. And all of that comes from a pretty high
level of spatial intelligence. But his other intelligences
were much lacking.
Id. at 68. On cross-examination, Dr. Draper supplied greater detail about
Mr. Harris’s history of intelligence testing:
[W]hen he was tested during his early school years it was
low/normal IQ, I believe, in the low 80s as a full scale. And it
was only later, after the fact, after the incident, I think he was
given a battery of tests by several different examiners and he was
found to have an IQ that ranged from the 60s to the 80s.
Id. at 133.
Both sides presented closing arguments on Mr. Harris’s intellectual
functioning. In their arguments, the prosecutors acknowledged that Mr.
Harris had a low IQ, but questioned the reliability of Dr. Draper’s
40
testimony about past IQ tests. The prosecutors also pointed out that
another expert witness had opined that Mr. Harris was malingering and told
the jury:
Who ever told you that he had a low IQ and that made it difficult
for him to solve problems? He can solve problems. He just
doesn’t solve them in a way that we think is appropriate. Jimmy
Dean Harris doesn’t have any problem with the way he solves
problems. It’s the rest of us that need to fear him for his problem-
solving abilities.
2005 Tr., v. 6, at 935.
The defense countered with Dr. Draper’s testimony. Defense counsel
urged the jury to focus on
[t]he images of a kid who falls behind in school because he just
can’t read. He’s got dyslexia, but he’s also close to mentally
retarded. We don’t have an exact number, but Dr. Draper
testified that 75 was the best consensus of all the numbers that
she looked at in the 60 hours that she prepared, talking to
everybody in this case, looking into his life.
Id. at 944. The attorney later emphasized Mr. Harris’s “75 IQ and real lack
of problem-solving skills,” noting that Dr. Draper had “talked about [how]
a person with a little better makeup, a little better development,” would
have been able to navigate the marital conflict without resorting to murder.
Id. at 960.
(b) Mitigation Evidence Involving an Intellectual Disability
On appeal, Mr. Harris argues that his trial counsel was ineffective in
failing to present mitigation evidence involving both an intellectual
disability and borderline intellectual functioning. But in the OCCA, Mr.
41
Harris did not argue that defense counsel should have presented mitigation
evidence involving an intellectual disability.
In briefing the issue to the OCCA, defense counsel was specific,
confining his argument to mitigation evidence involving borderline
intellectual functioning. In making this argument, defense counsel
considered intellectual disability an issue that could be addressed only in a
pretrial hearing. If the defendant prevailed, he would be ineligible for the
death penalty under Atkins v. Virginia, 536 U.S. 304, 321 (2002). If the
defendant lost on this issue, defense counsel apparently assumed that he
would have been barred from urging mitigation based on an intellectual
disability. Cf. Blonner v. State, 127 P.3d 1135, 1144 (Okla. Crim. App.
2006) (stating that if the pretrial hearing results in a finding of no
intellectual disability, “[t]he issue of mental retardation shall not be
relitigated at the capital first degree murder trial”). 22
Defense counsel thus acknowledged that he was not alleging a failure
to present additional mitigation evidence involving “mental retardation.”
He was instead confining the argument to additional evidence of a lesser
intellectual impairment that he called “borderline mental retardation,”
22
At oral argument, Mr. Harris contends the opposite, insisting that he
could have urged mitigation based on an intellectual disability even if the
state trial court had found no intellectual disability as a bar to execution.
But the OCCA did not have the benefit of this argument. In the OCCA, Mr.
Harris had disclaimed any argument that he could relitigate the existence
of an intellectual disability at the penalty phase.
42
presumably a synonym for Dr. Callahan’s preferred term “borderline
intellectual functioning”:
Appellant is not here claiming only to be borderline
mentally retarded – his claim of mental retardation is addressed
in Proposition I. However, given the procedural posture of this
case, counsel could not have argued that Mr. Harris was mentally
retarded since the mentally retarded are exempt from the death
penalty. If counsel had simply taken the previous testimony at
face value and not conducted an independent investigation into
Mr. Harris’ mental deficiencies, then he would have had
overwhelming evidence that Mr. Harris was borderline mentally
retarded. On the other hand, had defense counsel independently
investigated his client’s mental condition and determined that a
sufficient basis existed for a jury determination of the mental
retardation issue, it is likely that such a hearing would have been
held. In such a case, either Mr. Harris would have been
determined to be retarded, or not, by a jury. In this scenario,
counsel would have argued borderline mental retardation
because had a jury determined Mr. Harris to be mentally
retarded, then there would have been no capital sentencing at all.
Appellant’s Opening Br. at 16–17 n.15, No. D-2005-117 (Okla. Crim. App.
May 18, 2006).
Given Mr. Harris’s framing of the issue, the OCCA never referred to
an issue involving an intellectual disability. See Strelecki v. Okla. Tax
Comm’n, 872 P.2d 910, 925 n.1 (Okla. 1993), clarified on reh’g (Okla.
Mar. 23, 1994) (“[C]ourts are not free to act as advocates and to raise
claims that should be raised by the parties.”). The court instead referred to
“diminished mental capacity,” presumably as a synonym for defense
counsel’s term “borderline mental retardation” or Dr. Callahan’s preferred
term “borderline intellectual functioning.” So the OCCA addressed only
43
the lack of mitigation evidence involving borderline intellectual
functioning (not an intellectual disability).
Mr. Harris’s failure to present the OCCA with his current argument
would ordinarily constitute nonexhaustion of state-court remedies. See 28
U.S.C. § 2254(b)(1)(A). But exhaustion is unnecessary when it would be
futile. Selsor v. Workman, 644 F.3d 984, 1026 (10th Cir. 2011). And
exhaustion now would be futile because the OCCA would undoubtedly
consider the claim waived. See Slaughter v. State, 105 P.3d 832, 833
(Okla. Crim. App. 2005). 23 Mr. Harris’s claim is thus subject to a
procedural default, 24 and consideration of the merits would be available
only if Mr. Harris shows cause and prejudice. Banks v. Workman, 692 F.3d
1133, 1144 (10th Cir. 2012). Because Mr. Harris cannot show cause and
prejudice, we apply an anticipatory procedural bar and decline to consider
this claim. See Pavatt v. Carpenter, 928 F.3d 906, 924 (10th Cir. 2019) (en
banc) (holding that the habeas petitioner’s appellate argument was subject
23
Mr. Harris has already pursued a direct appeal and post-conviction
proceedings in which he could have (but failed to) raise this argument.
24
The State contends that even if the claim is unexhausted, the court
could deny relief on the merits under the AEDPA. It’s true that
unexhausted claims can be denied on the merits. 28 U.S.C. § 2254(b)(2).
But if the OCCA had not decided the claim on the merits, the AEDPA
would not apply. See pp. 3–4, 10, above.
44
to an anticipatory procedural bar because the argument had not been fairly
presented to the OCCA). 25
(c) Mitigation Evidence Involving Borderline Intellectual
Functioning
We also reject Mr. Harris’s claim that his counsel was ineffective in
presenting mitigation evidence on borderline intellectual functioning.
i. The OCCA’s Reliance on Both Prongs (Deficient
Performance and Prejudice)
On this claim, the OCCA concluded that Mr. Harris had not shown
either deficient performance or prejudice. Harris v. State, 164 P.3d 1103,
1116–18 (Okla. Crim. App. 2007). On the prong of deficient performance,
the court
noted that counsel had presented some evidence that involved
intellectual impairments,
discussed the virtually unchallengeable nature of strategic
decisions, and
concluded that defense “[c]ounsel’s choice of mitigating
evidence did not amount to ineffective assistance.”
25
Mr. Harris contends that the State failed to preserve its current
argument that defense counsel had not acted deficiently in failing to urge
mitigation based on an intellectual disability. But we ordinarily consider
an appellee’s arguments for affirmance even if they had not been presented
in district court. See United States v. Mosley, 743 F.3d 1317, 1324 & n.2
(10th Cir. 2014) (considering an argument for affirmance made by the
government for the first time on appeal even though the argument
conflicted with the government’s position in district court); see also
United States v. Bagley, 877 F.3d 1151, 1154 (10th Cir. 2017) (“Though
the government did not raise this argument in district court, we can affirm
on alternative grounds when the district court record is adequately
developed.”).
45
Id. at 1103, 1116, 1118. In this discussion, the OCCA rejected the claim at
least partly based on Mr. Harris’s failure to show a deficiency in the
representation.
On the prejudice prong, the OCCA referred to Mr. Harris’s argument
“that that the prejudice from this decision is evident.” Id. at 1118. The
OCCA rejected this argument, finding that the jurors at the retrial had
chosen the death sentence even after hearing some of this mitigating
evidence. Id.
ii. Deficient Performance
Mr. Harris claims that defense counsel should have presented
additional mitigation evidence on his borderline intellectual functioning.
For this claim, Mr. Harris argues that the OCCA made an unreasonable
determination of fact under § 2254(d)(2). According to Mr. Harris, the
OCCA unreasonably found that Mr. Harris’s attorney had strategically
chosen to bypass additional mitigation evidence. Mr. Harris argues that if
his attorney had conducted a reasonable investigation, he would have
learned of the evidence presented in the 2001 proceedings and would have
used a better expert witness to explain the evidence of borderline
intellectual functioning. The OCCA concluded that trial counsel’s
performance was neither deficient nor prejudicial. Harris v. State, 164
46
P.3d 1103, 1116–18 (Okla. Crim. App. 2007). These conclusions were
reasonable under § 2254(d)(2).
We begin with the OCCA’s determination that defense counsel’s
selection of evidence had been strategic. Mr. Harris argues that the OCCA
made an unreasonable factual determination because the state-court record
shows that defense counsel had not made a strategic decision. For this
argument, Mr. Harris states that
nothing in the record supported the OCCA’s determination that
Mr. Harris’s attorney had made a strategic decision and
after the penalty phase in the 2005 retrial, the attorney
continued to list Mr. Harris’s low IQ and inadequate problem-
solving skills as mitigating factors.
But we do not regard a factual finding as unreasonable if “‘[r]easonable
minds reviewing the record might disagree’ about the finding in question.”
Wood v. Allen, 558 U.S. 290, 301 (2010) (quoting Rice v. Collins, 546 U.S.
333, 341–42 (2006) (alteration in original)).
Reasonable minds could conclude that Mr. Harris’s attorney had
strategically decided how to present the evidence. For example, the record
indicated that the attorney was aware of the evidence that had been
presented in the state-court proceeding. In a colloquy with the judge, the
attorney said: “I’m not calling any shrinks, I’m not calling any
psychiatrists or all of the other people that testified last time.” 2005 Tr., v.
47
5, at 150. The OCCA could reasonably infer from this testimony that
defense counsel
had known of the evidence in the 2001 trial and
had deliberately declined to present additional evidence of
intellectual deficiencies.
See Wood v. Allen, 558 U.S. 290, 301–02 (2010) (holding that evidence
that counsel had known about omitted evidence and chosen not to present it
to a jury could “fairly be read to support” the state court’s judgment that
counsel had acted strategically). 26
In the alternative, Mr. Harris argues that even if the OCCA had
reasonably found that counsel acted strategically, this strategy would not
have involved a reasonable investigation. This argument fails because the
OCCA reasonably applied Supreme Court decisions in finding that defense
counsel had not performed deficiently. Wiggins v. Smith, 539 U.S. 510, 523
(2003).
To assess this argument, we consider the investigation underlying the
strategy. Wiggins v. Smith, 539 U.S. 510, 523 (2003). Mr. Harris argues
that the investigation was unreasonable because the attorney had
26
Mr. Harris also incorporates other arguments regarding an
unreasonable determination of fact. These arguments are addressed
elsewhere. For instance, Mr. Harris’s other arguments about the scope of
the investigation are better understood as arguments for reversal under 28
U.S.C. § 2254(d)(1); we thus consider these in our discussion of Mr.
Harris’s arguments under § 2254(d)(1).
48
known of evidence, presented in the 2001 trial, that Mr. Harris
was intellectually disabled and
engaged Dr. Draper (instead of another expert witness with
better qualifications) to discuss intellectual impairments.
The OCCA concluded that Mr. Harris’s attorney had decided not to
highlight the diagnoses and testing, choosing to focus instead on Mr.
Harris’s development throughout his life. Harris v. State, 164 P.3d 1103,
1118 (Okla. Crim. App. 2007). This conclusion was supported by the
record: Dr. Draper testified about Mr. Harris’s intellectual development
and his IQ testing. And in closing argument, defense counsel emphasized
Dr. Draper’s testimony about Mr. Harris’s overall development. Counsel
used that testimony to argue that an adult with greater development would
not have committed the murder.
This was not a case in which an attorney failed to investigate or
present any mitigation evidence on intellectual impairments. Rather, the
defense attorney pursued a strategy focusing on childhood development
rather than Mr. Harris’s mental state after the crime. And in implementing
this strategy, the attorney used a witness with expertise in personal
development. Applying the deferential AEDPA standard, we conclude that
defense counsel’s performance fell within the broad range of acceptable
strategies. See Doyle v. Dugger, 922 F.2d 646, 652 (11th Cir. 1991)
(concluding that defense counsel was not deficient for presenting only
some of the available evidence about the defendant’s mental state).
49
iii. Prejudice
Mr. Harris urges prejudice from his attorney’s failure to call an
expert on intellectual impairments, focusing on the “inherently mitigating”
nature of evidence of intellectual impairments when the death penalty is at
stake. Supp. Mem. Br. of Petitioner at 8 (quoting Tennard v. Dretke, 542
U.S. 274, 287 (2004)). The OCCA found no prejudice from defense
counsel’s failure to present additional mitigation evidence involving
borderline intellectual functioning. This finding was based on a reasonable
determination of facts and Supreme Court precedent. 27
We addressed an analogous issue in Grissom v. Carpenter, 902 F.3d
1265 (10th Cir. 2018). In Grissom, the petitioner claimed that his trial
attorneys had been ineffective by failing to investigate and present
evidence of organic brain damage because of “red flags” pointing to a
potentially fruitful defense on mitigation. 902 F.3d at 1272–73. We
affirmed the denial of habeas relief, explaining that the petitioner could
not show prejudice partly because his attorney had already presented a
27
We assume, for the sake of argument, that other evidence of
intellectual impairments would have been mitigating. In Tennard v. Dretke,
the Supreme Court recognized the inherently mitigating nature of evidence
involving intellectual impairments. 542 U.S. 274, 287 (2004). But in Atkins
v. Virginia, the Supreme Court noted that “reliance on mental retardation
as a mitigating factor can be a two-edged sword that may enhance the
likelihood that the aggravating factor of future dangerousness will be
found by the jury.” 536 U.S. 304, 321 (2002). That risk was arguably
present here because the State had alleged an aggravating circumstance of
future dangerousness.
50
robust mitigation case and the omitted report had “largely reflect[ed] the
mitigating narrative already presented at trial.” Id. at 1279 (quoting
Grissom v. State, 53 P.3d 969, 995 (Okla. Crim. App. 2011)).
This explanation is equally fitting here. Although a cognition expert
might have better emphasized the extent of an intellectual impairment,
defense counsel did not present the kind of “paradigmatic halfhearted
mitigation case” that we’ve regarded as constitutionally defective.
Littlejohn v. Royal, 875 F.3d 548, 563 (10th Cir. 2017), cert. denied, 139
S. Ct. 102 (2018). Instead, defense counsel presented seven fact witnesses
who testified about
Mr. Harris’s need for Pam’s help in reading technical
information, doing paperwork, and calling hotlines,
Mr. Harris’s difficulties in school because he was a slow
learner,
Mr. Harris’s dependable work,
verbal combat between Mr. Harris and Pam,
Pam’s berating of Mr. Harris,
childhood suffering of parental abuse, and
Mr. Harris’s loving relationship with his siblings and
daughters.
Defense counsel also presented Dr. Draper, who testified that Mr. Harris
was “slow,” had trouble in school, and needed help in working and
functioning in society. 2005 Tr., v. 5, at 58. Dr. Draper added that Mr.
51
Harris’s IQ scores were low, reflecting a high visual and spatial
intelligence that facilitated work as a transmission mechanic despite
shortcomings in other intellectual abilities. Id. at 68.
This testimony was not qualitatively different than Dr. Callahan’s
affidavit. Dr. Callahan assessed Mr. Harris’s intellectual status as
“borderline intellectual functioning.” R. at 289. And like Dr. Draper, Dr.
Callahan explained that Mr. Harris had strengths that allowed him to work
despite his intellectual deficits.
In closing argument, defense counsel also used Dr. Draper’s
testimony to emphasize Mr. Harris’s low intellectual ability and poor
problem-solving skills. Given the evidence and closing argument, the
OCCA could reasonably attribute little value to additional mitigation
evidence on borderline functioning. We thus conclude that the OCCA
reasonably applied Supreme Court precedents in finding no prejudice from
the failure to present greater evidence of borderline intellectual
functioning.
(d) Mitigation Evidence Involving Mental Illness 28
Mr. Harris also argues that his attorney was ineffective by failing to
call an expert witness specializing in mental health,
28
As noted above, we use the term “mental illness” to refer to various
cognitive and behavioral deficits not included in the other categories of
intellectual impairments (intellectual disability and borderline intellectual
functioning). See note 20, above.
52
highlight diagnoses of mental illness, and
show how mental illness might have contributed to the murder.
According to Mr. Harris, these shortcomings were prejudicial because the
additional evidence might have convinced at least one juror to vote for life
in prison rather than the death sentence. We reject this argument.
i. Mental-Health Evidence in the 2005 Retrial
At the 2005 retrial, defense counsel presented some evidence of
mental-health problems. But Mr. Harris argues that defense counsel should
have presented additional evidence from the 2001 trial and the competency
hearing.
At the 2005 retrial, defense counsel urged mitigation based on Mr.
Harris’s mental condition, alcoholism, drug abuse, and strong emotions.
But defense counsel did not call an expert witness specializing in mental
health; most of the evidence involving these mitigating factors came from
Dr. Draper.
Dr. Draper testified about three facets of mitigation:
1. When Mr. Harris had been a child, he suffered parental abuse
and saw his father abuse his mother.
2. As a teenager, Mr. Harris had obtained narcotics and alcohol
from his father, which led to a lifelong pattern of substance
abuse.
3. Mr. Harris had tried to commit suicide.
53
Dr. Draper added that eight to ten other doctors had found “serious
psychological problems”:
Q: You have been given various psychological tests that have
been administered to Jimmy Dean Harris over the years.
Have there been reports in there of any mental illnesses?
A: Well, there were. This was -- I think these were tests that
were administered after the incident.
Q: And approximately how many different doctors?
A: Well, there were probably eight or ten. I listed those in my
report; although, I did not pursue that area with any
diligence because that was after the fact. The significance
of that is all of those found that he had some serious
psychological problems.
2005 Tr., v. 5, at 67–68. Dr. Draper also testified that Mr. Harris was
taking medication to control his mental illness:
Q: And have you seen any records of medications given to him
in the jail?
A: Yes. I think he’s taking some psychotropic drugs and some
other medications for general health problems.
Q: And you’re not a physician, but you do know that the drugs
are for controlling mental illness?
A: Yes.
Id. at 68–69.
ii. Other Existing Evidence of Mr. Harris’s Mental Illness
In the prior proceedings, counsel for both parties elicited additional
evidence of Mr. Harris’s mental illness.
54
For example, Dr. John Smith testified that before the murder, Mr.
Harris had suffered from bipolar II disorder with psychosis. But Dr. Smith
conceded that it was difficult to pinpoint when Mr. Harris had experienced
the effects of drugs and alcohol.
An expert witness for the prosecution testified that Mr. Harris had
suffered from a major depressive episode with associated
psychotic features and
stabilized through medication.
And a jail counselor diagnosed Mr. Harris with schizo-affective disorder.
Dr. Smith and the jail counselor described Mr. Harris after the murder as
erratic, delusional, psychotic, and suicidal.
Other evidence suggested that Mr. Harris was responsive to
medication. Dr. Smith described these medications and opined that they
had helped, allowing Mr. Harris to attend the trial and testify with focus.
iii. Claim of Ineffective Assistance of Counsel
Mr. Harris argues that defense counsel was ineffective for failing to
call a mental-health expert and present this evidence in the penalty phase.
The OCCA rejected this claim without specifying whether the court was
relying on (1) the failure to show deficient representation or (2) prejudice.
Harris v. State, 164 P.3d 1103, 1116–19 (Okla. Crim. App. 2007). Given
55
this ambiguity, we apply 28 U.S.C. § 2254(d) on both prongs (deficient
performance and prejudice). Premo v. Moore, 562 U.S. 115, 123 (2011). 29
a. Deficiency Prong
Mr. Harris contends that the OCCA unreasonably determined the
facts and applied Supreme Court precedents.
(i) Unreasonable Factual Determinations
Mr. Harris argues that the OCCA based its decision on two
unreasonable determinations of fact:
1. that defense counsel had presented evidence of a mental illness
and
2. that defense counsel had strategically decided to downplay the
evidence of mental illness.
We reject both arguments.
Mr. Harris first points out that the OCCA said that “[w]hile Harris’s
specific diagnoses of mental illness [had not been] presented to the jury,”
jurors had been told that he was diagnosed as mentally ill. Harris v. State,
29
On this claim, Mr. Harris contends that the district court should have
conducted an evidentiary hearing. But the reasonableness of the OCCA’s
conclusion must be based on the existing state-court record. See Cullen v.
Pinholster, 563 U.S. 170, 181 (2011) (holding “that review under
§ 2254(d)(1) is limited to the record that was before the state court that
adjudicated the claim on the merits”); Hooks v. Workman, 689 F.3d 1148,
1163 (10th Cir. 2012) (stating that habeas review under § 2254(d)(2) is
also confined to the record in state court). Thus, the district court could
not consider evidence newly presented in federal court to determine
whether the OCCA had unreasonably applied federal law or determined the
facts.
56
164 P.3d 1103, 1118 (Okla. Crim. App. 2007). According to Mr. Harris,
this statement constituted an unreasonable determination of fact because
the State had denied the existence of any evidence of a mental illness.
We reject Mr. Harris’s argument. The OCCA observed that the jury
“had been told” of a diagnosis. Id. This observation was accurate, for Dr.
Draper had testified about a prior diagnosis of “serious psychological
problems.” 2005 Tr., v. 5, at 67–68; see p. 54, above. The OCCA’s
statement was thus reasonable based on the evidence presented.
The OCCA also stated that defense counsel had strategically decided
to downplay the evidence of mental illness. Mr. Harris argues that this
statement entailed an unreasonable factual determination. See 28 U.S.C.
§ 2254(d)(2). For this argument, he points to three facts:
1. Defense counsel asked Dr. Draper whether Mr. Harris’s
medications were for a mental illness, but defense counsel was
unable to obtain a response.
2. Defense counsel then tried to call an expert witness regarding
Mr. Harris’s medications, but the trial court sustained an
objection based on inadequate notice.
3. Despite the inability to obtain a response or call an expert
witness on medications, defense counsel continued to list Mr.
Harris’s “mental condition” as a mitigating factor.
Mr. Harris argues that these three facts show that defense counsel had tried
to prove a mental illness through an expert witness but couldn’t because
counsel had violated evidentiary and disclosure requirements.
57
A state appellate court’s finding may be reasonable even if we would
have decided the issue differently. Grant v. Trammell, 727 F.3d 1006, 1024
(10th Cir. 2013). The test is whether the state appellate court had
evidentiary support for its view. Id.
Under this test, the OCCA finding was reasonable. When defense
counsel called an expert witness to testify about Mr. Harris’s medications,
the judge asked the relevance. The attorney explained: “It goes to
mitigation, that he has something wrong with him and we don’t know what
it is.” 2005 Tr., v. 5, at 152. The attorney added that “[a]ll he is going to
do is say these are his medications and one, two, three, and four are mental
health medicines the other ones are for something else. And that’s all he is
going to say.” Id. at 149.
The OCCA could reasonably find that defense counsel was trying to
present limited evidence on mental health, informing the jury of a mental
illness without enough detail to spark concern about continued
dangerousness. The attorney couldn’t ultimately execute this strategy, but
the OCCA could view the strategy itself as reasonable. We thus conclude
that the OCCA acted reasonably in viewing defense counsel’s effort as
strategic.
(ii) Unreasonable Application of Supreme Court Precedents
The OCCA also reasonably applied Supreme Court precedents on the
deficiency prong. The ultimate failure of the attorney’s effort does not
58
undermine the reasonableness of the OCCA’s conclusion. United States v.
Haddock, 12 F.3d 950, 956 (10th Cir. 1993).
Mr. Harris contends that his attorney failed to present expert
testimony on the nature, extent, and significance of the mental illness.
Again, however, the OCCA acted reasonably in rejecting this contention.
As we discuss below, the excluded evidence constituted a “double-edged
sword” with substantial aggravating potential. See p. 62, below. Given this
potential for aggravation, the OCCA justifiably concluded that defense
counsel had acted reasonably.
b. Prejudice
Even if Mr. Harris could satisfy the deficiency prong, the claim
would have foundered on the prejudice prong. Mr. Harris argues that his
counsel’s failure to present mitigation evidence of mental illness (1)
opened the door to evidence of malingering and (2) bypassed powerful
mitigation evidence that would have explained Mr. Harris’s violent actions
and why, with proper treatment, he would be unlikely to repeat this crime.
In analyzing the prejudice prong, we consider not only the mitigation
evidence that defense counsel should have presented but also what the
prosecution would have presented in response. Wilson v. Trammell, 706
F.3d 1286, 1306 (10th Cir. 2013). To identify that evidence, we can
consider the 2001 trial as a useful guide. At that trial, the prosecution had
used Dr. John Call, who testified that
59
a “strong possibility” existed that Mr. Harris was a psychopath
and
psychopaths were more violent than other individuals.
2001 Tr., v. 16, at 75–78.
This testimony was supported by Mr. Harris’s own expert at the 2001
trial, Dr. Smith. Dr. Smith regarded Mr. Harris as bipolar and
acknowledged that bipolar individuals share traits with psychopaths. After
acknowledging the sharing of these traits, Dr. Smith refused to rule out Dr.
Call’s diagnosis of Mr. Harris as a psychopath or as someone with anti-
social personality disorder, admitting the presence of “elements” of these
conditions in Mr. Harris’s history and in his current psychological status.
2001 Tr., v. 18, at 182. Dr. Smith thus admitted that Mr. Harris presented a
substantial risk of violence. Id. at 183. On cross-examination, Dr. Smith
added that Mr. Harris had antisocial traits:
Q: But you don’t disagree with the diagnosis that [Mr. Harris]
has an antisocial personality disorder?
A: As long as you mix it with a mixed personality disorder
with narcissistic, obsessive-compulsive, and antisocial
traits. I do believe he does have that.
Id. at 192. Dr. Smith also acknowledged that Mr. Harris had each clinical
trait associated with antisocial personality disorder.
Given this prior testimony, the OCCA could reasonably conclude that
further mitigation testimony involving mental illness would have opened
the door to evidence of psychopathy with antisocial personality disorder.
60
“[C]ourts have characterized antisocial personality disorder as the
prosecution’s ‘strongest possible evidence in rebuttal.’” Littlejohn v.
Royal, 875 F.3d 548, 564 (10th Cir. 2017) (quoting Evans v. Sec’y, Dep’t
of Corr., 703 F.3d 1316, 1327 (11th Cir. 2013)), cert. denied, 139 S. Ct.
102 (2018).
We addressed a similar issue in Littlejohn v. Royal, 875 F.3d 548
(10th Cir. 2017), cert. denied, 139 S. Ct. 102 (2018). There the petitioner
presented evidence of organic brain disorder. The State responded with
evidence of a diagnosis involving antisocial personality disorder, and the
defense expert admitted that the petitioner had displayed traits consistent
with the diagnosis. Id. at 565. Given the nature of antisocial personality
disorder, we concluded that the evidence of an organic brain disorder was
likely to be aggravating rather than mitigating. Id.
The same is true here. Like organic brain damage, mental illness can
be mitigating; but the OCCA could reasonably view this possibility as
outweighed by the risk of rebuttal evidence of psychopathy and antisocial
personality disorder.
Mr. Harris argues that the jury at the 2005 retrial heard testimony
about his violent past with no explanation involving his mental illness. But
the OCCA could reasonably find that the aggravating nature of the omitted
evidence had outweighed the mitigation value.
61
By focusing on Mr. Harris’s development rather than his mental
illness, defense counsel also kept other possibly aggravating evidence from
the jury. For instance, the presence of an untreatable condition could have
suggested future dangerousness. Littlejohn, 875 F.3d at 565. And Dr. Smith
admitted that
he could not be confident that Mr. Harris would refrain from
violence while on medication,
Mr. Harris had probably been properly medicated during his
2001 competency trial (when he attacked a detention officer),
and
Mr. Harris had probably not been in a psychotic state when he
committed the murder.
With these admissions, Dr. Smith could not say whether Mr. Harris’s
mental illness was connected to the crime.
Finally, Mr. Harris argues that his counsel’s actions opened the door
to evidence of malingering. Even if defense counsel had presented
additional mental-health evidence, however, the State could still have
presented evidence of malingering. Indeed, at the 2001 trial, a prosecution
witness had testified that Mr. Harris was exaggerating the symptoms of any
mental illness. So the OCCA could reasonably consider evidence of
malingering as available irrespective of defense counsel’s strategy.
62
In sum, the OCCA acted reasonably in concluding that the omissions
were not prejudicial. 30
II. Jury Instructions and Closing Arguments on Mitigation Evidence
In capital cases, the Eighth and Fourteenth Amendments ordinarily
prevent the trial court from barring consideration of any of the defense
evidence on mitigation. Lockett v. Ohio, 438 U.S. 586, 604 (1978)
(plurality opinion). 31 Mr. Harris argues that the State violated this right in
two ways:
1. The trial court instructed the jury too narrowly on the evidence
that could be considered mitigating.
30
Mr. Harris points out that
the jury at the 2001 trial had declined to find the aggravator of
future dangerousness after hearing evidence of mental illness
and
the jury at the 2005 retrial did find this aggravator without
hearing that evidence.
But other possible explanations may account for the juries’ different
findings on future dangerousness. For example, the 2001 jury was
erroneously instructed on the availability of housing in a minimum-
security prison, and the jury at the 2005 retrial heard new evidence about
Mr. Harris’s violent actions. Given these differences, we decline to
speculate about why either jury found as it did.
31
The Lockett plurality stated:
[W]e conclude that the Eighth and Fourteenth Amendments
require that the sentencer, in all but the rarest kind of capital
case, not be precluded from considering, as a mitigating factor,
any aspect of a defendant’s character or record and any of the
circumstances of the offense that the defendant proffers as a
basis for a sentence less than death.
63
2. In closing argument, the first prosecutor exploited this
instruction by telling the jury that it should consider
mitigation evidence only if it diminished Mr. Harris’s moral
culpability.
These errors, according to Mr. Harris, created a reasonable likelihood that
one or more jurors believed themselves unable to consider some of Mr.
Harris’s mitigation evidence.
A. The Standard of Review
The OCCA rejected this claim. Because Mr. Harris did not object to
the instruction or the closing argument, the OCCA reviewed for plain
error. Harris v. State, 164 P.3d 1103, 1113 (Okla. Crim. App. 2007).
Applying the plain-error standard, the OCCA relied on its precedent to find
the jury instruction constitutional. E.g., Williams v. State, 22 P.3d 702, 727
(Okla. Crim. App. 2001), cited with approval in Harris, 164 P.3d at 1113
n.40. The OCCA thus focused on the prosecutors’ arguments, considering
how they might have affected the jury’s ability to consider mitigation
evidence. In considering this effect, the OCCA found that
the first prosecutor’s argument had been improper and
the second prosecutor’s argument and the jury instruction had
rendered the first prosecutor’s argument harmless.
Lockett, 438 U.S. at 604 (emphasis in original) (footnote omitted). A
majority of the Supreme Court later adopted this view in Eddings v.
Oklahoma, 455 U.S. 104, 113–15 (1982).
64
Harris, 164 P.3d at 1113–14.
We treat the OCCA’s decision under the plain-error standard as an
adjudication on the merits. Hancock v. Trammell, 798 F.3d 1002, 1011
(10th Cir. 2015); see pp. 3–4, above. We thus review this decision under
28 U.S.C. § 2254(d). Hancock, 798 F.3d at 1011–12; see pp. 3–4, above. 32
This review comprises two parts. We first ask “whether there is a
reasonable likelihood that the jury has applied the challenged instruction in
a way that prevents the consideration of constitutionally relevant
evidence.” Underwood v. Royal, 894 F.3d 1154, 1169 (10th Cir. 2018)
(quoting Boyde v. California, 494 U.S. 370, 380 (1990)), cert. denied, 139
S. Ct. 1342 (2019). We then ask whether a reasonable likelihood exists that
“arguments by the prosecutor . . . reinforced an impermissible
interpretation of [the challenged jury instruction] and made it likely that
jurors would arrive at such an understanding.” Id. (quoting Boyde, 494
U.S. at 384).
32
As noted above, the OCCA stated that the first prosecutor’s argument
was improper but harmless. Harris v. State, 164 P.3d 1103, 1113–14 (Okla.
Crim. App. 2007). But the test (discussed in the text) determines whether a
constitutional violation took place, not whether an error was harmless. See
Calderon v. Coleman, 525 U.S. 141, 146 (1998) (stating that Boyde’s test
of “reasonable likelihood” is used to determine whether a constitutional
error took place, not to determine harmlessness).
65
B. The Jury Instruction
Mr. Harris challenges Instruction Number 8, which stated in part:
“Mitigating circumstances are those which, in fairness, sympathy, and
mercy, may extenuate or reduce the degree of moral culpability or blame.”
R. at 1607. Mr. Harris argues that this instruction improperly prevented the
jury from considering all available mitigation evidence.
We rejected this argument in Hanson v. Sherrod, where we
considered the constitutionality of the same instruction. 797 F.3d 810 (10th
Cir. 2015). When faced with this argument, we addressed the instructions
as a whole. Id. at 851 (quoting Boyde v. California, 494 U.S. 370, 378
(1990)). Viewing them as a whole, we noted that three other jury
instructions had suggested that the jury would recognize its ability to
consider all of the defendant’s mitigation evidence:
1. The trial court had instructed the jury that it was to decide
which “circumstances [were] mitigating . . . under the facts and
circumstances of this case.”
2. Another jury instruction had identified many mitigating
circumstances, and some did not involve moral culpability.
3. The trial court had also instructed the jury that it “may decide
that other mitigating circumstances exist, and if so, [the jury]
should consider those circumstances as well.”
Id. Given these instructions, we concluded in Hanson that a jury would not
“have felt precluded from considering any mitigation evidence, including
the testimony of the four testifying witnesses.” Id.
66
The same three instructions were given here. So under Hanson, we
conclude that the OCCA reasonably determined that the jury would have
understood its ability to consider all of Mr. Harris’s mitigation evidence.
Hanson, 797 F.3d at 851; see Simpson v. Carpenter, 912 F.3d 542, 578
(10th Cir. 2018).
Mr. Harris argues that the OCCA has
expressed concern about the way that Oklahoma prosecutors
have used the jury instruction and
ordered revision of the jury instruction to minimize future
abuses.
Harris v. State, 164 P.3d 1103, 1114 (Okla. Crim. App. 2007). But we have
twice rejected the same argument, reasoning that the OCCA’s concern over
the wording of the instruction did not suggest that it was unconstitutional.
Hanson v. Sherrod, 797 F.3d 810, 851 (10th Cir. 2015); Grant v. Royal,
886 F.3d 874, 934–35 (10th Cir. 2018). Given these prior decisions, we
conclude that the OCCA’s concern over the instruction did not render its
constitutional holding unreasonable.
C. The Prosecutors’ Closing Arguments
Mr. Harris argues that even if the jury instruction itself had been
constitutional, one of the prosecutors improperly exploited the jury
instruction to urge disregard of Mr. Harris’s mitigation evidence, violating
the Eighth and Fourteenth Amendments. The OCCA rejected this argument.
The court acknowledged that the first prosecutor’s arguments had been
67
improper; however, the court considered the impropriety harmless because
the jury instructions on mitigating circumstances were proper and the
second prosecutor had invited the jury to consider all of the mitigating
circumstances. Harris v. State, 164 P.3d 1103, 1113 (Okla. Crim. App.
2007). 33 In the OCCA’s view, the “second prosecutor invited jurors to
consider all Harris’s mitigation evidence, weigh it against the aggravating
circumstances, and find that the death penalty was appropriate.” Id. Mr.
Harris contends that the OCCA acted unreasonably in finding facts and
applying Supreme Court precedents. We disagree because
the OCCA could reasonably view this part of the closing
argument as an invitation to consider all of the evidence on
33
At oral argument, the State defends the first prosecutor’s arguments,
stating that they invited the jury to consider mitigation evidence and to
give it little weight. For example, the first prosecutor acknowledged that
the jury
could consider “sympathy or sentiment for the defendant” and
needed to determine the importance of the mitigating
circumstances.
2005 Tr., v. 6, at 909. And this prosecutor acknowledged the jury’s need to
balance the mitigating and aggravating circumstances. Id. at 940–41. But
the OCCA found that the first prosecutor’s arguments had been improper:
One prosecutor did consistently argue in closing that jurors
should not consider Harris’s second stage evidence as
mitigating, since it did not extenuate or reduce his guilt or moral
culpability. This argument improperly told jurors not to consider
Harris’s mitigating evidence.
Harris v. State, 164 P.3d 1103, 1113 (Okla. Crim. App. 2007).
.
68
mitigation and find it overridden by the horrific nature of the
crime and
Mr. Harris has not shown that the OCCA based its decision on
an erroneous interpretation of the prosecutor’s closing
argument.
1. Applicability of 28 U.S.C. § 2254(d)
The threshold issue is whether § 2254(d) applies. It ordinarily would
apply if the OCCA adjudicated the merits of Mr. Harris’s constitutional
claim. See pp. 3–4, 10, above. The OCCA wasn’t explicit. It characterized
the first prosecutor’s closing argument as improper, but didn’t say whether
this impropriety rose to the level of a constitutional violation. Regardless
of the basis for characterizing the argument as improper, the OCCA
ultimately regarded the impropriety as harmless. Harris, 164 P.3d at 1113–
14.
The district court characterized the OCCA’s reasoning as an
adjudication on the merits, triggering § 2254(d). D. Ct. Dkt. 77 at 49. In
his appeal briefs, Mr. Harris doesn’t question this characterization. We
thus decline to sua sponte revisit the district court’s application of
§ 2254(d). See Grant v. Royal, 886 F.3d 874, 931–32 n.20 (10th Cir. 2018)
(concluding that the Court should not sua sponte reject the applicability of
the AEDPA on a claim involving the prosecutor’s improper exploitation of
a jury instruction defining the proper use of mitigating evidence).
69
2. Unreasonable Determination of Fact
In his rebuttal argument, the second prosecutor told the jury:
I’m asking you to make a decision that I believe is based upon
principal [sic], to examine the evidence, determine whether you
believe beyond a reasonable doubt one or both of these
aggravators are in existence, and I submit to you, and then to
make a determination of whether these—these mitigation issues
that [Mr. Harris’s attorney] has brought up really override the day
of terror, and a day that took a couple of weeks to think through.
2005 Tr., v. 6, at 982–83. In our view, the OCCA could reasonably
construe this statement as an invitation to weigh all of the mitigation
evidence against the aggravation evidence and decide that the death
sentence was appropriate.
The prosecutor did contend that the defense’s arguments on
mitigation would not “override the day of terror.” The term “override”
refers to the act of weighing one item against another. See Override,
Oxford English Reference Dictionary 1038 (Judy Pearsall & Bill Trumble
eds., 2d ed. rev. 2006) (providing a primary definition of “override” as
“have a claim precedence or superiority over”). Given this meaning of
“override,” the OCCA could reasonably conclude that the second
prosecutor had urged the jury to consider all of the mitigation evidence and
to find that it paled in comparison to the terrible nature of the crime itself.
Under this interpretation, the second prosecutor’s rebuttal argument would
not have restricted the universe of circumstances that could be considered
70
mitigating. Grant v. Royal, 886 F.3d 874, 938 (10th Cir. 2018), cert.
denied, 139 S. Ct. 925 (2019).
Mr. Harris notes that the second prosecutor also referred to moral
culpability: “Do not reward this man for the things that he claims are
somehow supposed to not make this as blameful, if you will, these things
that he says somehow lessen his blame, lessen his moral responsibility.”
2005 Tr., v. 6, at 983. According to Mr. Harris, this statement reflects
further efforts to restrict mitigating circumstances to those bearing on
moral culpability. This argument bears defects that are both procedural and
substantive.
The argument is procedurally defective because
in the state-court appeal, defense counsel never criticized the
second prosecutor’s reference to moral culpability and
in our appeal, defense counsel did not criticize this statement
in their opening brief.
Defense counsel instead referred to this excerpt only in their reply brief,
when the State no longer had an opportunity to respond. Making the
argument in the reply brief was too late. See Byrd v. Workman, 645 F.3d
1159, 1166 n.8 (10th Cir. 2011).
Even if we were to consider the second prosecutor’s reference to
moral culpability, however, it would not render the OCCA’s interpretation
unreasonable. The second prosecutor was responding to what defense
counsel had argued. Defense counsel had argued that Mr. Harris was not a
71
cold-blooded terrorist and was reacting to setbacks involving divorce and
unemployment. The second prosecutor characterized this argument as an
effort to minimize blame and moral culpability. With this characterization,
the prosecutor attributed the statement about blame and moral culpability
to defense counsel, arguing that “the [defense counsel] says [these things]
somehow lessen [Mr. Harris’s] blame, lessen his moral responsibility.”
2005 Tr., v. 6, at 983. The prosecutor himself was not suggesting that the
universe of mitigating circumstances should be limited to those that
diminish blame or moral culpability; he was saying that defense counsel’s
argument involved an effort to downplay blame and moral culpability. We
thus do not regard the OCCA’s interpretation of the rebuttal argument as
objectively unreasonable.
Even if the OCCA had unreasonably interpreted the rebuttal
argument, however, § 2254(d)(2) would prevent the district court from
reaching the merits. Under § 2254(d)(2), the habeas court can consider the
merits only if the petitioner shows that the OCCA based its decision on the
factual error. 28 U.S.C. § 2254(d)(2); see Lott v. Trammell, 705 F.3d 1167,
1177 (10th Cir. 2003) (stating that “the burden rests on [the petitioner] to
establish that the OCCA’s analysis was ‘based on an unreasonable
determination of the facts’” (quoting 28 U.S.C. § 2254(d)(2))).
In deciding that the first prosecutor’s improper arguments were
harmless, the OCCA gave two reasons:
72
1. The second prosecutor told the jury to consider all of Mr.
Harris’s mitigating circumstances and find that the death
penalty was appropriate based on the greater strength of the
aggravating circumstances.
2. The trial court properly instructed the jury on the definition of
mitigation evidence, Mr. Harris’s evidence, and the jurors’
duties.
Harris v. State, 164 P.3d 1103, 1113 (Okla. Crim. App. 2007).
Mr. Harris challenges the first reason, but not the second. Let’s
assume, for the sake of argument, that the first reason was objectively
unreasonable. Given this assumption, the OCCA’s second reason would
remain valid and provide sufficient support for the OCCA’s finding of
harmlessness:
[E]ven if a state court’s individualized factual determinations
are overturned, what factual findings remain to support the state
court decision must still be weighed under the overarching
standard of section 2254(d)(2).
Lambert v. Blackwell, 387 F.3d 210, 235–36 (3d Cir. 2004). Indeed, the
OCCA has elsewhere found improper closing arguments harmless when the
jury was properly instructed. E.g., Miller v. State, 313 P.3d 934, 977
(Okla. Crim. App. 2013); Ake v. State, 663 P.2d 1, 9 (Okla. Crim. App.
1983), rev’d on other grounds, 470 U.S. 68 (1985). So even if the OCCA
had unreasonably interpreted the second prosecutor’s closing argument,
Mr. Harris would have failed to show that the decision itself had been
based on this factual error.
* * *
73
In sum, Mr. Harris has not satisfied his burden of showing that the
OCCA based its decision on an unreasonable factual determination. The
OCCA could reasonably interpret the second prosecutor’s argument as an
invitation to consider all of the mitigation evidence and find it overridden
by the aggravating circumstances. The second prosecutor did mention
moral culpability, but Mr. Harris did not address this statement in the
state-court appeal or in his opening appellate brief. And the second
prosecutor referred to moral culpability only when he paraphrased defense
counsel’s argument. In these circumstances, Mr. Harris has not overcome
§ 2254(d)(2).
3. Unreasonable Application of Supreme Court Precedent
Mr. Harris also argues that the OCCA’s decision entailed an
unreasonable application of Supreme Court precedent. See 28 U.S.C.
§ 2254(d)(1). We reject this argument.
Like Mr. Harris and the OCCA, we view the first prosecutor’s
comments as improper. The first prosecutor told the jury that a mitigating
circumstance was something that “extenuates or reduces the degree of
moral culpability or blame of [Mr.] Harris for murdering Merle Taylor.”
2005 Tr., v. 6, at 929. The prosecutor then pointed to each alleged
mitigating circumstance and asked if it reduced or extenuated Mr. Harris’s
74
moral culpability. Id. at 929–40. The prosecutor then proposed a two-part
test:
One, is it true? Is what they have listed here true? Did it really
happen? And, two, if it is true, does it make a difference? Does
it extenuate or reduce his culpability for the murder of Merle
Taylor? Because it’s got to be both.
Id. at 930 (emphasis added). Through these statements, the prosecutor
effectively told the jury that the mitigation evidence mattered only if it
tended to reduce Mr. Harris’s culpability, creating a risk that one or more
jurors believed that they could not consider constitutionally relevant
evidence of mitigation. See Underwood v. Royal, 894 F.3d 1154, 1169
(10th Cir. 2018), cert. denied, 139 S. Ct. 1342 (2019).
We thus inquire whether “the OCCA could reasonably conclude that
it was not reasonably likely that the [first] prosecutor’s comment[s]
precluded the jury from considering mitigation evidence, in light of the
jury instructions and the other unchallenged comments of the prosecution.”
Grant v. Royal, 886 F.3d 874, 939 (10th Cir. 2018), cert. denied, 139 S.
Ct. 925 (2019). Under this inquiry, a court could grant habeas relief only if
“no fairminded jurist would agree with the OCCA’s conclusion that the
jury was not precluded from considering the evidence offered by [the
petitioner] in mitigation.” Simpson v. Carpenter, 912 F.3d 542, 582 (10th
Cir. 2018). In our view, fair-minded jurists could have agreed with the
75
OCCA’s conclusion in light of the jury instructions and the second
prosecutor’s rebuttal argument.
When the petitioner argues that a prosecutor exploited a jury
instruction to improperly restrict what could be mitigating, we consider the
extent to which the jury was properly instructed. See, e.g., Grant, 886 F.3d
at 939; Hanson v. Sherrod, 797 F.3d 810, 852 (10th Cir. 2015). The jury at
the 2005 retrial received virtually all of the jury instructions that we have
regarded as curative. For example, the trial court instructed the jury that
“the determination of what circumstances are mitigating is for
you to resolve under the facts and circumstances of this case”
and
evidence had been introduced on a long list of mitigating
circumstances (many of which bore no relationship to moral
culpability).
See pp. 66–67, above. In detailing the mitigating circumstances, the trial
court reminded the jury of evidence that Mr. Harris
had a “sister and a brother who love him” and “daughters who
love[d] and need[ed] him,”
had a “low I.Q.,”
had been addicted to drugs and alcohol, and
had lost his mother to cancer when he was young.
R. at 1608–10. These instructions served to broaden the first prosecutor’s
language, suggesting to the jury that it could consider all of the mitigation
evidence regardless of whether it related to moral culpability. Hanson, 797
76
F.3d at 851; see Brown v. Payton, 544 U.S. 133, 144 (2005) (“[F]or the
jury to have believed it could not consider Payton’s mitigating evidence, it
would have had to believe that the penalty phase served virtually no
purpose at all.”).
In similar circumstances, we have often held that prosecutors’
improper arguments on mitigation evidence are ameliorated by the jury
instructions. E.g., Grant v. Royal, 886 F.3d 874, 939–42 (10th Cir. 2018);
Underwood v. Royal, 894 F.3d 1154, 1171–73 (10th Cir. 2018); Simpson v.
Carpenter, 912 F.3d 542, 581–82 (10th Cir. 2018); Cuesta-Rodriguez v.
Carpenter, 916 F.3d 885, 911–12 (10th Cir. 2019); Johnson v. Carpenter,
918 F.3d 895, 907–08 (10th Cir. 2019); Harmon v. Carpenter, 936 F.3d
1044, 1074–77 (10th Cir. 2019). For example, in Cuesta-Rodriguez v.
Carpenter, the trial court instructed the jury on numerous mitigating
circumstances, told the jury that it was to determine what was mitigating,
and stated to the jury that it could consider sympathy for the defendant.
Cuesta-Rodriguez, 916 F.3d at 911–12. Given these instructions, we held
that the OCCA had reasonably applied Supreme Court decisions in
rejecting a similar constitutional claim. Id. at 912. All of these instructions
were given here.
Mr. Harris contends that in one of our prior cases, Grant v. Royal,
the trial court had given two instructions that were omitted here:
77
1. that the jury instructions contained all of the law and rules for
the jury to follow and
2. that the prosecutor’s closing arguments were arguments only
and for purposes of persuasion.
We reject these contentions.
Mr. Harris contends that the Grant panel found it “critically
ameliorative” that the trial court had told the jury that the instructions
contained all of the law and rules to be followed. Appellant’s Reply Br. at
32. Though the Grant panel did consider this instruction, along with
others, the panel did not suggest that this instruction was “critical” to the
outcome. Instead, the Grant panel simply mentioned this instruction “[i]n
addition” to others. Grant, 886 F.3d at 941. Indeed, many of our opinions
recognize the ameliorative impact of other jury instructions with no
indication that the jury had been told that the instructions constituted all of
the law and rules to be followed. E.g., Simpson, 912 F.3d at 581–82;
Cuesta-Rodriguez, 916 F.3d at 911–12; Johnson, 918 F.3d at 907–08.
But even if this instruction had been critical, it was given to Mr.
Harris’s jury. Just after voir dire, the trial court instructed Mr. Harris’s
jury that its responsibility was “to follow the law as stated in the
instruction that [the trial court] will give [the jury].” 2005 Tr., v. 2, at 427.
The trial court returned to the subject later, explaining what would likely
happen if the jury were to ask questions during its deliberations. 2005 Tr.,
v. 6, at 984. The court explained that it would likely answer that the jury
78
has “all the law and evidence necessary to reach a verdict.” Id. The court
explained that this answer would mean that all of the necessary
information is in the jury instructions or the evidence. Id. at 984–85. Thus,
Mr. Harris’s jury was ultimately told that all of the applicable law was in
the instructions.
Mr. Harris also observes that the jury in Grant had been told that the
prosecutor’s remarks constituted only argument and were offered only for
persuasion. 886 F.3d at 941–42. Mr. Harris says that this instruction was
“critical” in Grant. Appellant’s Reply Br. at 32. We are not sure why Mr.
Harris regards this instruction as critical, for the Grant panel attached no
particular importance to this instruction. In any event, Mr. Harris’s jury
was instructed to confine itself to the evidence and reminded that “[n]o
statement or argument of the attorneys [was] itself evidence.” 2005 Tr., v.
2, at 428.
Along with the ameliorating jury instructions, some of the second
prosecutor’s arguments also mitigated the risk from the first prosecutor’s
improper arguments. For example, the second prosecutor told the jury to
weigh the defense’s evidence against the aggravating evidence to see if the
mitigation evidence outweighed the aggravating evidence. 34 And both
34
Mr. Harris insists that the second prosecutor did not suggest to the
jury that it consider any of the mitigation evidence. According to Mr.
Harris, the absence of such a suggestion distinguishes Simpson v.
Carpenter and Grant v. Royal. We disagree. As noted above, the OCCA
79
prosecutors spent considerable time rebutting the defense’s mitigation
evidence even when it had not involved moral culpability. The prosecutors
attacked this evidence not only because it bore no relationship to moral
culpability but also on grounds that the evidence lacked reliability or
trustworthiness. For example, the first prosecutor attacked the reliability of
Mr. Harris’s evidence on an intellectual impairment. From this attack, the
jury could “logically infer from this presentation that the evidence actually
did legally qualify as mitigating evidence, and that the question before
them” involved the accuracy, credibility, and weight of this evidence.
Grant v. Royal, 886 F.3d 874, 943 (10th Cir. 2018) (emphasis in original).
Mr. Harris underscores the repeated nature of the first prosecutor’s
improper comments. But we’ve upheld the reasonableness of a similar
conclusion by the OCCA even when the prosecutor had made at least “nine
separate statements which either generally defined mitigating evidence as
reasonably concluded that the second prosecutor had invited the jury to
consider all of the evidence, both mitigating and aggravating. See pp. 71–
74, above. But this factor was not present in Simpson. There we described
the prosecutor’s improper arguments as “pervasive,” “extensive,” and
“recurring.” 912 F.3d 542, 581, 588 (10th Cir. 2018), petition for cert.
filed (U.S. July 24, 2019) (No. 19-5298). Nowhere did we rely on
arguments inviting the jury to weigh the mitigation evidence. See id. at
585–87. The same is true in Grant. 886 F.3d 874, 943 (10th Cir. 2018)
(“To be sure, unlike Hanson, there were no further statements from the
prosecution — i.e., Ms. Elliott — in rebuttal closing that could reasonably
suggest that ‘the prosecutor encouraged the jury to consider all sorts of
mitigating evidence.’” (quoting Hanson v. Sherrod, 797 F.3d 810, 852
(10th Cir. 2015))).
80
reducing moral culpability or blame or specifically compared [the
petitioner’s] mitigating factors to that definition.” Simpson v. Carpenter,
912 F.3d 542, 578 (10th Cir. 2018). And there the prosecutor had not said
anything to encourage consideration of all mitigating factors. Id. at 580;
see note 34, above.
Given the ameliorating jury instructions and the closing arguments as
a whole, fair-minded jurists could agree with the OCCA’s conclusion that
the jury had understood its ability to consider Mr. Harris’s mitigation
evidence. We thus conclude that the OCCA did not unreasonably apply
Supreme Court precedent.
III. Victim-Impact Testimony
Mr. Harris also contends that the prosecution improperly elicited
victim-impact testimony. Though some of the testimony was
unconstitutional, the constitutional violation was harmless.
A. The Constitutional Limit on Victim-Impact Testimony
Mr. Harris’s contention stems from the interplay between two
Supreme Court opinions: Booth v. Maryland and Payne v. Tennessee. In
Booth v. Maryland, the Supreme Court held that the introduction of victim-
impact testimony at a capital-sentencing proceeding violated the
Constitution. 482 U.S. 496, 509 (1987). In Payne v. Tennessee, the
Supreme Court overruled part of Booth, holding that “evidence and
argument relating to the victim and the impact of the victim’s death on the
81
victim’s family are []admissible at a capital sentencing hearing.” 501 U.S.
808, 830 n.2 (1991). But the Payne Court did not overrule Booth’s
recognition that the Constitution forbids “the admission of a victim’s
family members’ characterizations and opinions about the crime, the
defendant, and the appropriate sentence.” Id. Thus, Booth continues to ban
the families of murder victims from requesting a particular sentence. Bosse
v. Oklahoma, 580 U.S. ___, 137 S. Ct. 1, 2 (2016) (per curiam).
B. The Victim-Impact Testimony and the Issue of Harmlessness
In Mr. Harris’s case, two of Mr. Taylor’s family members requested
the death penalty. Mr. Harris argues that allowing this testimony violated
the Constitution. The OCCA rejected this argument. Harris v. State, 164
P.3d 1103, 1110 (Okla. Crim. App. 2007). The OCCA was wrong:
Introduction of this testimony was unconstitutional under Booth and Payne,
and the OCCA’s decision was contrary to clearly established Supreme
Court precedent. Dodd v. Trammell, 753 F.3d 971, 996 (10th Cir. 2013).
The remaining question is whether the constitutional error was
prejudicial or harmless. On this question, we engage in de novo review.
Lockett v. Trammell, 711 F.3d 1218, 1238 (10th Cir. 2013). We regard the
improper testimony as prejudicial only if it had a substantial and injurious
effect or influence in determining the jury’s verdict. Id.
Mr. Taylor’s son testified for the State, asking for the death penalty:
“On behalf of myself, my entire family, I respectfully ask that you impose
82
the maximum allowable punishment and, in my mind, the only acceptable
punishment, and sentence [Mr.] Harris to death.” 2005 Tr., v. 4, at 891.
Mr. Taylor’s widow also testified, asking the jury to impose the death
penalty: “It grieves me that my husband went to his grave not knowing why
he had to die. My sons, grandchildren, and I ask you to sentence [Mr.]
Harris to death.” Id. at 901.
In her closing argument, the first prosecutor did not refer to the
family members’ requests for the death penalty. She instead urged the jury:
“Do not be guilted into making your decision because . . . the Taylors are
going to be upset, frankly. Make your decision because it is right, it is just,
it is what is appropriate.” 2005 Tr., v. 6, at 935. Similarly, the second
prosecutor did not explicitly mention the family members’ requests for the
death penalty. But this prosecutor did quote extensively from the family
members’ testimony, urging the jury not to reward Mr. Harris by sparing
his life. Right after asking the jury one more time not to “reward [Harris],”
the second prosecutor continued, “Toby Taylor [the son] and Carolyn
Taylor [the widow] said this.” Id. at 979. The prosecutor then summarized
the family members’ testimony on how they were affected by the murder.
C. Structural or Harmless Error
The threshold issue is whether a habeas court can review for
harmlessness when the trial court improperly allows victim-impact
83
testimony. Mr. Harris opposes review for harmlessness and urges us to
treat the requests for the death penalty as structural error, contending that
Oklahoma prosecutors regularly elicit family requests for the
death penalty and
the OCCA has improperly tolerated this pattern of improper
conduct. 35
But we have rejected the same arguments in Underwood v. Royal, holding
that erroneous introduction of victim-impact testimony is reviewable for
harmlessness. 894 F.3d 1154, 1177 (10th Cir. 2018). We are bound by this
precedent. Leatherwood v. Albaugh, 861 F.3d 1034, 1042 n.6 (10th Cir.
2017). Given this precedent, we consider whether the error was harmless. 36
35
Until 2017, the Oklahoma Court of Appeals had interpreted Payne to
overrule Booth in its entirety. Bosse v. State, 360 P.3d 1203, 1226 (Okla.
Crim. App. 2015). The Supreme Court expressly rejected the OCCA’s view
in Bosse v. State, reiterating that Payne had left intact Booth’s prohibition
against a family member’s request for a particular sentence. Bosse v.
Oklahoma, 580 U.S. ___, 137 S. Ct. 1, 2 (2016) (per curiam). On remand,
the OCCA overruled its prior cases and held that the Constitution forbids
victim-impact testimony recommending a particular sentence. Bosse v.
State, 400 P.3d 834, 855 (Okla. Crim. App. 2017).
36
The Supreme Court has noted that an unusual case might involve a
pattern of prosecutorial misconduct so egregious that habeas relief might
be appropriate even without prejudice. Brecht v. Abrahamson, 507 U.S.
619, 638 n.9 (1993). But when the 2005 retrial took place in an Oklahoma
courtroom, Oklahoma’s highest criminal court had held that the
Constitution did not forbid victim testimony requesting a particular
sentence. E.g., Murphy v. State, 47 P.3d 876, 885 (Okla. Crim. App. 2002),
overruled in part on other grounds by Blonner v. State, 127 P.3d 1135,
1139 (Okla. Crim. App. 2006). We had said the opposite. Hooper v. Mullin,
314 F.3d 1162, 1174 (10th Cir. 2002). But Oklahoma prosecutors were
simply following what Oklahoma’s highest criminal court had said on the
issue.
84
D. Harmlessness
We regard the erroneous introduction of victim-impact testimony as
harmless.
For harmlessness, we consider whether the constitutional error “had
substantial and injurious effect or influence in determining the jury’s
verdict.” Brecht v. Abrahamson, 507 U.S. 619, 623 (1993) (quoting
Kotteakos v. United States, 328 U.S. 750, 776 (1946)). On one occasion,
we concluded that improper victim-impact testimony had a substantial and
injurious effect or influence. Dodd v. Trammell, 753 F.3d 971, 997 (10th
Cir. 2013). There we relied on three factors:
1. The prosecution had elicited a “drumbeat” consisting of
six to seven witnesses requesting the death penalty.
2. The jury had rejected the State’s arguments for
aggravating circumstances involving a “heinous,
atrocious, or cruel” murder or the existence of a
“continuing threat.”
3. The case for the defendant’s guilt had not been clear-cut.
Even if we were to regard the prosecutor’s conduct as egregious,
however, a habeas court could avoid the issue of harmlessness only if the
victim-impact testimony had rendered the trial fundamentally unfair.
Underwood v. Royal, 894 F.3d 1154, 1178 (10th Cir. 2018). As discussed
elsewhere, the improper testimony consisted of two sentences in a five-day
trial. Though the two sentences were emotional and powerful, they did not
render the entire trial fundamentally unfair.
85
Id. at 997–98. None of these factors are present here. Only two testifying
witnesses requested death, far from a “drumbeat.” 37 The jury also found the
aggravator of a continuing threat, and Mr. Harris has not challenged his
guilt.
Other factors also point to harmlessness, including the ameliorating
influence of the jury instructions, the brevity of the improper testimony,
and the absence of any mention in the prosecutors’ closing arguments. For
example, the trial court instructed the jury that it could consider the
evidence “in determining an appropriate punishment,” but only as “a moral
inquiry into the culpability of the defendant” and not based on an
“emotional response to the evidence.” R. at 1616. The jury was also told
that it could consider “sympathy or sentiment for the defendant.” Id. at
1618 (emphasis in original). These instructions mitigated the prejudicial
impact of the improper victim-impact testimony. DeRosa v. Workman, 679
F.3d 1196, 1240 (10th Cir. 2012). We consider not only the ameliorating
instructions but also the brevity of the improper testimony, which
37
Even when the prosecution presents a “drumbeat” of improper victim
testimony, the constitutional violation may be harmless. In Bush v.
Carpenter, for example, “sentence recommendations were lengthy [and]
egregious.” 926 F.3d 644, 668 (10th Cir. 2019); see also id. at 680 (“[T]he
victim impact statements were numerous, emotional, and in at least one
instance, egregious . . . .”). Still, we held that the constitutional violation
was harmless “given the circumstances of the murder, the presence of the
aggravating factors, and the substantial evidence presented in support of
those aggravating factors.” Id. at 681.
86
consisted of only two sentences. See Lockett v. Trammell, 711 F.3d 1218,
1239 (10th Cir. 2013) (considering the error to be harmless when the
family’s requests for death consisted of “a single, concise sentence”). 38
And in their closing arguments, the prosecutors did not explicitly refer to
the family members’ requests for the death penalty.
Mr. Harris argues that the State presented a weak case on
aggravation. 39 We disagree. The jury found two aggravators:
1. great risk of death to more than one person and
2. continuing threat.
Mr. Harris does not challenge the sufficiency of the evidence on either
aggravator, and the State presented powerful evidence on both.
First, to show a great risk of death to more than one person, the State
presented evidence that Mr. Harris had not only killed Mr. Taylor but also
38
Mr. Harris argues that the son’s request was expansive, consisting of
seventeen pages of argument on why the death penalty was the only
appropriate punishment. But the son’s testimony mainly concerned the
effect of the crime, which was permissible. See Payne v. Tennessee, 501
U.S. 808, 827 (1991) (“A State may legitimately conclude that evidence
about the victim and about the impact of the murder on the victim’s family
is relevant to the jury’s decision as to whether or not the death penalty
should be imposed.”).
39
For this argument, Mr. Harris relies on Dodd v. Trammell, 753 F.3d
971, 998 (10th Cir. 2013), where we discounted the aggravating factors
because they had added little beyond the findings of guilt. Dodd, 753 F.3d
at 998. There the jury’s finding of an aggravator involving a prior
conviction had been based on a decades-old conviction, and the aggravator
for great risk of death to more than one person had been based on the fact
that the defendant had murdered two people. Id.
87
fired multiple times at Pam Harris and Jennifer Taylor. Pam Harris
testified that she had suffered a gunshot to her hip and had seen the gun
aimed at her head. She struggled as Mr. Harris tried to reload the gun,
which he then used to smash her on the head and face.
Second, the State presented considerable evidence of the aggravator
involving a continuing threat. This evidence included
bar fights,
physical abuse of Pam Harris,
intimidating tactics, and
threats against Pam Harris’s family.
Given this evidence, the OCCA reasonably found “a lifelong pattern of
using violence to solve problems and react to situations which is likely to
continue.” Harris v. State, 164 P.3d 1103, 1111 (Okla. Crim. App. 2007).
Mr. Harris, of course, would have been imprisoned for life if he had
avoided the death penalty. But even while he was in jail, Mr. Harris had
assaulted a guard. In this incident, Mr. Harris covered his cell window and
surprised the guard, repeatedly pummeling him.
Mr. Harris attributes this assault to his need for medication. But Dr.
Smith acknowledged that Mr. Harris had probably been medicated at the
time of the assault. 40
40
Mr. Harris suggests that county officers might have “messed up” his
medications, stating that the Oklahoma Department of Corrections is much
88
* * *
We conclude that the constitutional error did not substantially affect
the jury’s sentencing recommendation, so the district court acted correctly
in rejecting this habeas claim.
IV. Cumulative Error
Mr. Harris also urges cumulative error. In our view, the district court
should revisit this issue on remand.
A cumulative-error analysis aggregates all errors that are
individually harmless, analyzing whether the cumulative effect undermines
confidence in the fairness of the retrial and reliability of the
verdict. Workman v. Mullin, 342 F.3d 1100, 1116 (10th Cir. 2003). We
consider cumulative errors to be separate constitutional violations. Hanson
v. Sherrod, 797 F.3d 810, 852 n.16 (10th Cir. 2015).
When we reject a claim of ineffective assistance based on a lack of
prejudice, we can aggregate the prejudice from the deficient performance.
Cargle v. Mullin, 317 F.3d 1196, 1207 (10th Cir. 2003). As a result, the
claim of cumulative error would ordinarily include the prejudice from two
claims:
more reliable in administering medication. Appellant’s Reply Br. at 23. For
this suggestion, however, Mr. Harris relies on evidence from the 2001 trial,
not the 2005 retrial involved in this appeal. In the 2005 retrial, no one
presented evidence of an error in medicating Mr. Harris before this assault.
89
1. any prejudice from counsel’s failure to seek a pretrial hearing
on an intellectual disability and
2. an error in admitting the victim-impact testimony.
On the claim of cumulative error, the OCCA also included any
incremental prejudice from the first prosecutor’s closing argument about
the jury’s consideration of mitigation evidence. Harris v. State, 164 P.3d
1103, 1119 (Okla. Crim. App. 2007). We have held that Mr. Harris failed
to show an unreasonable legal or factual determination on the
constitutionality of the closing arguments. See pp. 67–81, above. Though
we have not recognized a constitutional violation involving the closing
arguments, the constitutional test bears a close resemblance to the test for
harmlessness. See Boyde v. California, 494 U.S. 370, 393 (1990)
(Marshall, J., dissenting) (“[T]he ‘reasonable likelihood’ standard should
be understood to be an equivalent of the ‘harmless error’ standard adopted
in Chapman v. California.”). Arguably, then, any incremental prejudice
from this claim may need to be combined with the prejudice from defense
counsel’s failure to seek a pretrial hearing on an intellectual disability and
a constitutional error in allowing the victim-impact testimony.
But the parties have not briefed whether this claim should be
considered in the mix on the claim of cumulative error. We thus leave
consideration of this threshold issue to the district court on remand. See
Greystone Const., Inc. v. Nat’l Fire & Marine Ins. Co., 661 F.3d 1272,
90
1290 (10th Cir. 2011) (“[T]he better practice on issues raised [below] but
not ruled on by the district court is to leave the matter to the district court
in the first instance.” (quoting Apartment Inv. & Mgmt. Co. v. Nutmeg Ins.
Co., 593 F.3d 1188, 1198 (10th Cir. 2010))).
The State also contends that in analyzing the claim of cumulative
error, the court should not include any prejudice from the failure to request
a pretrial hearing on an intellectual disability, asserting that the prejudice
would have arisen before the trial and could not “accumulate with trial
errors.” Appellee’s Resp. Br. at 96. But all we have are two sentences
without any explanation, authority, or response. So we also leave this
second threshold issue for the district court to decide in the first instance.
See Greystone Const., Inc., 661 F.3d at 1290.
Motion to Expand the Certificate of Appealability
Mr. Harris moves to expand the certificate of appealability to include
whether “trial counsel breached his duty to Mr. Harris by his failure to
present as mitigation a psychological risk assessment to diminish the
evidence presented by the State that Mr. Harris posed a continuing threat
to society.” Appellant’s Mot. for Modification of Certificate of
Appealability at 2 (text case changed).
At the 2005 retrial, the State urged an aggravating circumstance
involving Mr. Harris’s continued threat. The defense countered with Dr.
Draper, who testified that
91
Mr. Harris had been incarcerated for over 1800 days with only
one incident,
Mr. Harris would not be dangerous in the structured
environment of a prison,
the availability of proper medication would remove any
possible danger, and
murderers are generally less likely than others to act violently
while in prison.
Mr. Harris argues that defense counsel should have presented expert
testimony of a risk assessment. In state court, for example, Mr. Harris
presented a risk assessment by J. Randall Price, Ph.D. The OCCA rejected
this argument, concluding that defense counsel had acted reasonably at the
2005 retrial. Harris v. State, 164 P.3d 1103, 1118–19 (Okla. Crim. App.
2007).
We could grant a certificate of appealability on this issue only if the
district court’s ruling were debatable among reasonable jurists. Miller-El v.
Cockrell, 537 U.S. 322, 338 (2003). Because the OCCA adjudicated the
merits of the deficiency prong, the federal district court would need to
apply § 2254(d) on this prong. See pp. 55–56, above. Mr. Harris could thus
obtain a certificate of appealability on this claim only by showing that
reasonable jurists could debate his ability to clear the hurdle of § 2254(d).
See Dunn v. Madison, 583 U.S. ___, 138 S. Ct. 9, 11 (2017) (per curiam).
No reasonable jurist would regard this issue as debatable. As the
OCCA noted, defense counsel
92
had countered the prosecution with Dr. Draper, who testified
that Mr. Harris would not pose a significant risk of future
violence in a structured environment, and
had strategic reasons to limit the evidence of future
dangerousness.
Mr. Harris contends that Dr. Price could have provided more
persuasive evidence. But Dr. Price’s opinion created two risks:
1. His opinion could have backfired.
2. Dr. Price had diagnosed Mr. Harris as bipolar with psychotic
features, which could have led to further evidence of
dangerousness.
Dr. Price opined that even in a maximum-security prison, Mr. Harris
had “an 18.8% probability of violent conduct.” Appl. for Evid. Hearing,
Exh. B-2 at 7. In stating this opinion, Dr. Price defined “violent” conduct
as an “assaultive or dangerous” act creating an imminent threat of serious
bodily injury. Id. at 6. The OCCA could reasonably infer that defense
counsel might have regarded an 18.8% risk of future violence as high.
Indeed, Dr. Price acknowledged that this percentage exceeded the base rate
for capital murderers (16.4%). Id. at 7.
Second, Dr. Price noted that Mr. Harris had “been diagnosed as
bipolar with psychotic features.” Id. at 8. An acknowledgment of psychotic
features could have led the State to present additional evidence of future
dangerousness. In the 2001 trial, for example, Dr. Smith acknowledged that
bipolar disorder and psychopathy share many of the same characteristics.
93
2001 Tr., v. 18, at 179–80. This sort of testimony in the 2005 retrial could
have been “devastating.” See United States v. Barrett, 797 F.3d 1207, 1232
(10th Cir. 2015).
Given the possibility that a risk assessment might backfire, defense
counsel could reasonably focus instead on Mr. Harris’s difficult upbringing
and on his generally positive conduct while in prison. See Lott v.
Trammell, 705 F.3d 1167, 1209 (10th Cir. 2013) (stating that defense
counsel was not ineffective for failing to present a risk assessment because
cross-examination could have yielded negative information increasing the
chances for a death sentence). “In fact, counsel would have been
ineffective if the door to the damaging Risk Assessment Report and
evidence contained therein had been opened and the State had been able to
exploit it to their advantage.” Id. We thus deny Mr. Harris’s motion to
expand the certificate of appealability.
Conclusion
We reverse on Mr. Harris’s claim of ineffective assistance in defense
counsel’s failure to seek a pretrial hearing on an intellectual disability. On
remand, the district court should revisit the issue of prejudice after
conducting an evidentiary hearing.
We also vacate the district court’s judgment on the claim of
cumulative error. On this claim, the district court should first consider the
threshold issues of whether it can consider the prejudice arising from
94
the lack of a request for a pretrial hearing on intellectual
disability and
the first prosecutor’s exploitation of the jury instruction on Mr.
Harris’s mitigation evidence.
On the claim of cumulative error, the court should also consider the
prejudice resulting from the constitutional error in allowing victim-impact
testimony recommending the death penalty.
We affirm the district court’s ruling in all other respects and deny
Mr. Harris’s motion to expand the certificate of appealability.
95