United States Court of Appeals
FOR THE DISTRICT OF COLUMBIA CIRCUIT
Argued February 22, 2016 Decided June 10, 2016
No. 14-3060
UNITED STATES OF AMERICA,
APPELLEE
v.
QUEEN NWOYE,
APPELLANT
Appeal from the United States District Court
for the District of Columbia
(No. 1:07-cr-00012-1)
A.J. Kramer, Federal Public Defender, argued the cause
and filed the briefs for appellant.
Karen P. Seifert, Assistant U.S. Attorney, argued the
cause for appellee. On the brief were Elizabeth Trosman,
Frederick Yette, James Sweeney, and David P. Saybolt,
Assistant U.S. Attorneys.
Before: KAVANAUGH, Circuit Judge, and EDWARDS and
SENTELLE, Senior Circuit Judges.
2
Opinion for the Court filed by Circuit Judge
KAVANAUGH, with whom Senior Circuit Judge EDWARDS
joins.
Dissenting opinion filed by Senior Circuit Judge
SENTELLE.
KAVANAUGH, Circuit Judge: A woman named Queen
Nwoye was convicted of conspiring with her boyfriend,
Adriane Osuagwu, to extort money from a prominent doctor
with whom Nwoye had previously had an affair. At trial,
Nwoye testified that she acted under duress: She said that
Osuagwu repeatedly beat her and forced her to participate in
the extortion scheme. Despite asserting a duress defense
based on Osuagwu’s repeated abuse of Nwoye, Nwoye’s
counsel did not introduce expert testimony on battered woman
syndrome. At the close of trial, Nwoye’s counsel requested a
jury instruction on duress, but the District Court denied the
request. A jury then convicted Nwoye of conspiracy to
commit extortion.
On direct appeal, Nwoye challenged the District Court’s
failure to instruct the jury on duress. This Court rejected the
challenge, with Judge Tatel dissenting. But the Court
indicated that it was rejecting Nwoye’s challenge in part
because Nwoye had failed to introduce expert testimony on
battered woman syndrome. Such expert testimony, the Court
suggested, may have entitled Nwoye to a duress instruction.
Nwoye then moved to vacate the conviction based on
alleged ineffective assistance of trial counsel. A claim of
ineffective assistance of counsel requires the defendant to
show (i) that counsel’s performance was constitutionally
deficient and (ii) that the ineffective assistance prejudiced the
defendant. On the first prong, the deficiency prong, Nwoye
claimed that competent trial counsel would have introduced
3
expert testimony on battered woman syndrome. On the
second prong, the prejudice prong, Nwoye claimed that such
expert testimony would have led the District Court to instruct
the jury on duress. And Nwoye further argued that the
combination of the expert testimony and the duress instruction
would have created a reasonable doubt respecting her guilt.
The District Court denied Nwoye’s ineffective-
assistance-of-counsel claim. The District Court held that
Nwoye was not prejudiced by trial counsel’s failure to
introduce expert testimony on battered woman syndrome.
The District Court therefore did not need to (and did not)
decide whether counsel’s performance was constitutionally
deficient.
Although the prejudice question is close, we see it
differently than the District Court did. We conclude that
Nwoye was prejudiced by trial counsel’s failure to introduce
expert testimony on battered woman syndrome. We therefore
reverse the judgment of the District Court, and we remand so
that the District Court may decide whether Nwoye’s counsel
was constitutionally deficient in failing to present such
testimony.
I
A
In January 2007, a woman named Queen Nwoye was
indicted for conspiring with her then-boyfriend, Adriane
Osuagwu, to extort money from Ikemba Iweala. Iweala was a
prominent doctor. He and Nwoye had previously had an
affair. Over the course of 49 days in 2006, Osuagwu and
Nwoye repeatedly threatened Iweala that they would
publicize his prior relationship with Nwoye unless Iweala
paid them. Their threats were effective. Iweala made six
4
separate payments to Osuagwu and Nwoye, totaling almost
$200,000.
At Nwoye’s trial, Nwoye admitted to engaging in the
alleged extortion but testified that Osuagwu had coerced her
participation through his physically abusive and controlling
behavior. According to Nwoye, her relationship with
Osuagwu turned abusive shortly after they started dating in
2005. Osuagwu would frequently slap Nwoye with his hand,
hit her with his shoe, and beat her on her face and body.
Later, Osuagwu’s physical violence escalated. Osuagwu beat
Nwoye when she initially refused to introduce him to Iweala.
Whenever she objected to the extortion, Osuagwu would beat
her “like a drum.” And on one occasion when Nwoye did not
play her part in the extortion scheme, Osuagwu slapped
Nwoye and threatened to “strangle” and “kill” her if the
scheme were exposed.
Nwoye further testified that Osuagwu exerted financial
and psychological control over her. Osuagwu forced Nwoye
to hand over her ATM card and PIN. In addition, Nwoye and
her children lived with Osuagwu at Osuagwu’s home in
Maryland. Nwoye testified that Osuagwu – the only person
who knew that she lived at the house – would often threaten
to kill Nwoye and bury her inside the house. Nwoye also
testified that she was afraid to report Osuagwu to the police
because Osuagwu had told her that he was a former FBI
agent.
At the same time, Nwoye’s testimony revealed that
Osuagwu did not have direct physical control over Nwoye at
all times. While Nwoye attended nursing school or worked at
a hospital for three days a week, she was apart from
Osuagwu. And Osuagwu spent at least a few days in
California while Nwoye remained in Maryland.
5
But even while they were apart, Osuagwu constantly
monitored Nwoye. He forced Nwoye to keep her phone with
her and demanded that she answer promptly, even going so
far as to require Nwoye to wear a Bluetooth earpiece during
class at nursing school.
B
Despite the significant evidence of Nwoye’s abusive
relationship with Osuagwu, Nwoye’s trial counsel did not
seek to introduce expert testimony on battered woman
syndrome.
Battered woman syndrome is a term that was coined by
Dr. Lenore Walker in the late 1970s to describe the
psychological and behavioral traits common to women who
are exposed to severe, repeated domestic abuse. See LENORE
E. WALKER, THE BATTERED WOMAN SYNDROME (1984);
LENORE E. WALKER, THE BATTERED WOMAN (1979). Dr.
Walker’s theory was that women subject to cyclical domestic
abuse develop psychological paralysis – or “learned
helplessness” – that renders them unable to escape abusive
relationships. See WALKER, BATTERED WOMAN SYNDROME
at 86-97.1
1
Although the majority of domestic violence victims are
women, some cases involve victims who are men. See Department
of Justice, Bureau of Justice Statistics, Intimate Partner Violence,
1993-2010 1 (Nov. 2012, rev. Sept. 2015) (“From 1994 to 2010,
about 4 in 5 victims of intimate partner violence were female.”).
Some scholars have advocated abandoning the term “battered
woman syndrome” in favor of the label “battering and its effects.”
See, e.g., Mary Ann Dutton, Update of the “Battered Woman
Syndrome” Critique, National Online Resource Center on Violence
Against Women (Aug. 2009). We use the term “battered woman
syndrome” in this opinion because the term is commonly used by
6
Since the advent of Dr. Walker’s influential research,
courts have admitted expert testimony on battered woman
syndrome to support claims of duress and self-defense. See
Janet Parrish, Trend Analysis: Expert Testimony on Battering
and Its Effects in Criminal Cases, in DEPARTMENT OF JUSTICE
ET AL., THE VALIDITY AND USE OF EVIDENCE CONCERNING
BATTERING AND ITS EFFECTS IN CRIMINAL TRIALS pt. II, at 19,
21-22, 28 (1996) (hereinafter DOJ REPORT).
C
At Nwoye’s trial, Nwoye’s counsel did not present expert
testimony on battered woman syndrome. Counsel instead
staked Nwoye’s duress defense entirely on Nwoye’s own trial
testimony. At the close of trial, Nwoye’s counsel requested a
jury instruction on duress. To be entitled to an instruction on
duress, Nwoye had to present sufficient evidence (i) that she
acted under an unlawful threat of imminent death or serious
bodily injury and (ii) that there was no reasonable alternative
to participating in the extortion scheme. See United States v.
Jenrette, 744 F.2d 817, 820-21 (D.C. Cir. 1984).
The District Court ruled that Nwoye had not presented
sufficient evidence on the second prong of duress – the no-
reasonable-alternative prong – and therefore declined to give
the duress instruction. The jury then convicted Nwoye of
conspiracy to commit extortion, and the District Court
sentenced Nwoye to 20 months in prison, followed by three
years of supervised release.2
courts and because it describes the alleged circumstances in this
case.
2
For his part, Osuagwu pled guilty to conspiracy and was
sentenced to 22 months in prison.
7
Nwoye appealed, challenging the District Court’s
decision not to give the duress instruction. This Court
affirmed. United States v. Nwoye, 663 F.3d 460 (D.C. Cir.
2011). The Court explained that Nwoye had a number of
reasonable alternatives to participating in the extortion
scheme, including reporting Osuagwu to police or to friends
and co-workers when she was at school or work, away from
Osuagwu. Id. at 462-65.
The Court also stressed that although Nwoye had testified
about the abuse she suffered, she failed to present “other usual
indicia supporting a BWS defense – expert witnesses
testifying to the effects of isolation, financial dependence, or
estrangement from family members.” Id. at 465. Therefore,
the Court concluded that Nwoye was not entitled to a jury
instruction on duress.
Judge Tatel dissented. In his view, Nwoye’s testimony
concerning Osuagwu’s threats and abuse amounted to “more
than enough evidence to have warranted a duress instruction.”
Id. at 468.
D
In 2013, after the termination of her supervised release,
Nwoye filed a motion to vacate her conviction.3 Nwoye
3
On appeal, Nwoye has characterized her post-conviction
motion as a petition for a writ of coram nobis. “A petition for a
writ of coram nobis provides a way to collaterally attack a criminal
conviction for a person . . . who is no longer ‘in custody’ and
therefore cannot seek habeas relief under 28 U.S.C. § 2255 or
§ 2241.” Chaidez v. United States, 133 S. Ct. 1103, 1106 n.1
(2013). The Government does not dispute that a petition for a writ
of coram nobis is an appropriate way for Nwoye to seek redress for
8
claimed that her trial counsel was constitutionally ineffective
because counsel failed to call an expert witness to testify
about battered woman syndrome. To establish that her
counsel was constitutionally ineffective, Nwoye had to prove
(i) that her counsel’s performance was constitutionally
deficient and (ii) that counsel’s ineffective assistance
prejudiced her. Strickland v. Washington, 466 U.S. 668, 687
(1984).
The District Court held an evidentiary hearing to allow
Nwoye to present the expert testimony on battered woman
syndrome that Nwoye argued should have been offered at
trial. At the hearing, Nwoye’s expert – Dr. Carole Giunta –
testified extensively about battered woman syndrome in
general. Dr. Giunta also opined that Nwoye’s relationship
with Osuagwu exhibited the “classic dynamics” of a battering
relationship.
After considering this new evidence, the District Court
denied Nwoye’s motion to vacate her conviction. United
States v. Nwoye, 60 F. Supp. 3d 225 (D.D.C. 2014). Applying
the two-part test for ineffective-assistance claims, the District
Court ultimately did not decide whether trial counsel’s alleged
failures rose to the level of a constitutional deficiency.
Instead, the District Court decided that Nwoye was, in any
event, not prejudiced by the lack of expert testimony on
battered woman syndrome. The District Court reasoned that
such testimony still would not have satisfied the second prong
of duress – the no-reasonable-alternative prong – and that the
testimony therefore would not have entitled Nwoye to a jury
instruction on duress. Id. at 236-42. For that reason, the
her claim of ineffective assistance of counsel. See United States v.
Newman, 805 F.3d 1143, 1146 (D.C. Cir. 2015).
9
District Court held that Nwoye had not made out a successful
claim of ineffective assistance of counsel.
II
Nwoye claims that her trial counsel was constitutionally
ineffective because counsel failed to present expert testimony
on battered woman syndrome. To reiterate, a defendant
claiming ineffective assistance of counsel must show that (i)
“counsel’s representation fell below an objective standard of
reasonableness” and (ii) “the deficient performance
prejudiced the defense.” Strickland v. Washington, 466 U.S.
668, 687-88 (1984). The District Court held that trial
counsel’s failure to present expert testimony on battered
woman syndrome was not prejudicial. The District Court thus
did not need to (and did not) decide whether trial counsel’s
failure to present such testimony was constitutionally
deficient.
On appeal, therefore, the only issue for us to decide is
whether the failure of Nwoye’s trial counsel to present expert
testimony on battered woman syndrome was prejudicial. If
so, then we must remand for the District Court to determine
whether trial counsel was constitutionally deficient in failing
to introduce such testimony. If not, then we must affirm. We
review de novo the District Court’s denial of Nwoye’s claim
of ineffective assistance of counsel, including the question
whether Nwoye was prejudiced by her counsel’s allegedly
deficient performance. See United States v. Abney, 812 F.3d
1079, 1086-87 (D.C. Cir. 2016).4 Although the dissent
4
Of course, we still review for clear error any findings of
historical fact embedded in the District Court’s conclusions on
deficient performance and prejudice. See Payne v. Stansberry, 760
F.3d 10, 13 (D.C. Cir. 2014); see also HARRY T. EDWARDS, LINDA
A. ELLIOTT, & MARIN K. LEVY, FEDERAL STANDARDS OF REVIEW:
10
suggests that we should give some deference to the District
Court, our case law is clear that the standard of review is de
novo.
To establish prejudice, Nwoye must demonstrate “a
reasonable probability that, absent the errors, the factfinder
would have had a reasonable doubt respecting guilt.”
Strickland, 466 U.S. at 695. To demonstrate a reasonable
probability, Nwoye “need not show that counsel’s deficient
conduct more likely than not altered the outcome in the case.”
Id. at 693. She must demonstrate only “a probability
sufficient to undermine confidence” in the verdict. Id. at 694.
Nwoye’s prejudice argument proceeds in two steps:
First, Nwoye argues that expert testimony on battered woman
syndrome would have entitled her to a jury instruction on the
defense of duress. Second, she claims that a duress
instruction, together with the expert testimony on battered
woman syndrome, would have created a “reasonable
probability” that the jury “would have had a reasonable doubt
respecting guilt.” Id. at 695. Nwoye must succeed on both
arguments to establish prejudice.
We conclude first that expert testimony on battered
woman syndrome would have entitled Nwoye to a jury
instruction on duress. And we conclude further that a duress
instruction, together with the expert testimony, would have
created a reasonable probability that the jury would have had
a reasonable doubt respecting Nwoye’s guilt.
REVIEW OF DISTRICT COURT DECISIONS AND AGENCY ACTIONS 1
(2d ed. Supp. 2015) (When “courts determine that a particular
mixed question of law and fact should be treated as a question of
law and reviewed de novo, subsidiary findings of fact are properly
reviewed under the clearly erroneous standard of review.”).
11
A
As this Court stated on Nwoye’s direct appeal, a
defendant is entitled to a jury instruction on any recognized
affirmative defense “if there is sufficient evidence from which
a reasonable jury could find for the defendant on that theory.”
United States v. Nwoye, 663 F.3d 460, 462 (D.C. Cir. 2011)
(internal quotation marks omitted). To obtain an instruction
on the affirmative defense of duress, a defendant must
produce sufficient evidence (i) that “she acted under an
unlawful threat of imminent death or serious bodily injury,”
and (ii) that there was no “reasonable, legal alternative to
committing the crime.” Id. (internal quotation marks
omitted).
On direct appeal, this Court held that the evidence
Nwoye presented at trial was insufficient to support a duress
instruction. Id. at 462-65. In so ruling, however, the Court
emphasized that Nwoye had not introduced expert testimony
on battered woman syndrome. Id. at 465. The Court
indicated (or at least implied) that Nwoye might have been
entitled to a duress instruction had she introduced such
testimony. We now must decide that question.
The question, put simply, is whether expert testimony on
battered woman syndrome would have moved the evidentiary
needle enough to entitle Nwoye to a duress instruction. To
answer that question, we must initially assess whether, in
general, expert testimony on battered woman syndrome can
be admissible to prove duress – that is, whether it can be
reliable and can be relevant to the duress defense. If so, then
we next must assess whether the particular expert testimony
proffered by Nwoye in her post-conviction proceeding was
reliable and would have provided relevant evidence at
Nwoye’s trial. Finally, if Nwoye’s expert testimony would
12
have been admissible, we must determine whether the
introduction of such testimony at Nwoye’s trial would have
entitled her to a jury instruction on duress.
We now address those three questions in turn.
1
For expert testimony to be admissible in federal court, it
must be both reliable and relevant. See FED. R. EVID. 702;
Daubert v. Merrell Dow Pharmaceuticals, Inc., 509 U.S. 579,
589 (1993).
As to reliability, trial judges possess “considerable
leeway in deciding in a particular case how to go about
determining whether particular expert testimony is reliable.”
Kumho Tire Co. v. Carmichael, 526 U.S. 137, 152 (1999). At
bottom, though, the expert’s opinion must have “a reliable
basis in the knowledge and experience of his discipline.” Id.
at 148.
The Government does not dispute that expert testimony
on battered woman syndrome can be reliable. And several
courts of appeals have permitted the admission of expert
testimony on battered woman syndrome. See United States v.
Young, 316 F.3d 649, 656-59 (7th Cir. 2002); Arcoren v.
United States, 929 F.2d 1235, 1241 (8th Cir. 1991). None has
held that such testimony is categorically unreliable, so far as
we are aware. Given the history of expert testimony on this
subject and the extensive literature, we too agree that expert
testimony on battered woman syndrome can be reliable,
assuming of course that the expert can demonstrate sufficient
expertise to meet the usual requirements for experts to testify
on a subject.
13
To be admissible in support of a duress defense, expert
testimony on battered woman syndrome must also be relevant
to proving duress. Most courts that have considered the
question – especially in recent years – have recognized that
expert testimony on battered woman syndrome can be
relevant to prove duress. See Dando v. Yukins, 461 F.3d 791,
801 (6th Cir. 2006); United States v. Dixon, 413 F.3d 520,
524 & n.3 (5th Cir. 2005); United States v. Ceballos, 593 F.
Supp. 2d 1054, 1060-62 (S.D. Iowa 2009); United States v.
Marenghi, 893 F. Supp. 85, 91-97 (D. Me. 1995); Wonnum v.
State, 942 A.2d 569, 572-73 (Del. 2007); Commonwealth v.
Pike, 726 N.E.2d 940, 948 (Mass. 2000) (expert testimony
assumed to be relevant to prove duress); People v. Romero, 13
Cal. Rptr. 2d 332 (1992), rev’d on other grounds, 883 P.2d
388 (Cal. 1994); but see United States v. Willis, 38 F.3d 170,
174-77 (5th Cir. 1994); State v. B.H., 870 A.2d 273, 289-91
(N.J. 2005) (expert testimony irrelevant to reasonableness of
duress defense).
We agree with the majority of the courts that expert
testimony on battered woman syndrome can be relevant to the
duress defense. The reason, put simply, is that the duress
defense requires a defendant to have acted reasonably under
the circumstances, and expert testimony can help a jury assess
whether a battered woman’s actions were reasonable.
Reasonableness is the touchstone of a duress defense. To
satisfy the first prong of the duress defense, the defendant
must have acted under the influence of a reasonable fear of
imminent death or serious bodily harm at the time of the
alleged crime. See 2 WAYNE R. LAFAVE, SUBSTANTIVE
CRIMINAL LAW § 9.7(b) (2d ed. 2003) (“the danger need not
be real; it is enough if the defendant reasonably believes it to
be real”); see also United States v. Jenrette, 744 F.2d 817,
820-21 (D.C. Cir. 1984). And to satisfy the second prong of
14
the defense, there must have been no “reasonable, legal
alternative to committing the crime.” Nwoye, 663 F.3d at 462
(emphasis added). Whether an alternative is reasonable turns
on whether a reasonable person would have availed herself of
it.
Reasonableness – under both the imminence prong and
the no-reasonable-alternative prong – is not assessed in the
abstract. Rather, any assessment of the reasonableness of a
defendant’s actions must take into account the defendant’s
“particular circumstances,” at least to a certain extent. See id.
at 464; see also Model Penal Code § 2.09 (duress defense
appropriate whenever a “person of reasonable firmness in his
situation would have been unable to resist” threat of unlawful
force) (emphasis added).
The circumstances that juries have historically considered
in assessing reasonableness have been factors “that
differentiate the actor from another, like his size, strength,
age, or health,” as well as facts known to the defendant at the
time in question, such as the defendant’s knowledge of an
assailant’s violent reputation. Model Penal Code § 2.09 cmt.
at 375 (1985); Smith v. United States, 161 U.S. 85, 88 (1896).
On the other hand, courts have typically precluded juries from
considering factors such as the defendant’s particular
“psychological incapacity” or her “clarity of judgment,
suggestibility or moral insight.” Model Penal Code § 2.09
cmt. at 373-74 (1985).
Thus, whether expert testimony on battered woman
syndrome is relevant to the duress defense turns on whether
such testimony can identify any aspects of the defendant’s
“particular circumstances” that can help the jury assess the
reasonableness of her actions. Examination of the particulars
of the duress defense shows that expert testimony on battered
15
woman syndrome can indeed identify relevant aspects of a
battered woman’s particular circumstances.
With respect to the first prong of the duress defense – the
imminent-harm prong – women in battering relationships are
often “hypervigilant to cues of impending danger and
accurately perceive the seriousness of the situation before
another person who had not been repeatedly abused might
recognize the danger.” Lenore E.A. Walker, Battered Women
Syndrome and Self-Defense, 6 NOTRE DAME J.L. ETHICS &
PUB. POL’Y 321, 324 (1992). Remarks or gestures that may
seem harmless to the average observer might be reasonably
understood to presage imminent and severe violence when
viewed against the backdrop of the batterer’s particular
pattern of violence. As our colleague Judge Brown stated
while on the California Supreme Court: “Although a jury
might not find the appearances sufficient to provoke a
reasonable person’s fear, they might conclude otherwise as to
a reasonable person’s perception of the reality when
enlightened by expert testimony on the concept of
hypervigilance.” People v. Humphrey, 921 P.2d 1, 17 (Cal.
1996) (Brown, J., concurring).
Regarding the second prong of the duress defense – the
no-reasonable-alternative prong – battered women face
significant impediments to leaving abusive relationships.
Most importantly, battered women who leave their abusers
risk a retaliatory escalation in violence against themselves or
those close to them – sometimes termed “separation abuse.”
Mary Ann Dutton, Validity of “Battered Woman Syndrome”
in Criminal Cases Involving Battered Women, in DOJ REPORT
pt. I, at 14-15; Desmond Ellis, Post-Separation Woman
Abuse: The Contribution of Lawyers as “Barracudas,”
“Advocates,” and “Counsellors,” 10 INT’L J.L. &
PSYCHIATRY 403, 408 (1987). For example, studies have
16
suggested that women in battering relationships are more
likely to be killed by their batterers after separating from
them. See Dutton, Validity of “Battered Woman Syndrome”
in Criminal Cases Involving Battered Women (citing Margo
Wilson et al., Uxoricide in Canada: Demographic Risk
Patterns, 35 CANADIAN J. CRIMINOLOGY 263, 263-91 (1993)),
in DOJ REPORT pt. I, at 14. In addition, batterers often isolate
their victims and exert financial control over them, rendering
separation a significant burden. See LENORE E. WALKER, THE
BATTERED WOMAN, 129-32 (1979). Expert testimony on
those impediments to separation can help explain why a
battered woman did not take advantage of an otherwise
reasonable-sounding opportunity to avoid committing the
alleged crime.
In short, expert testimony on battered woman syndrome
can be relevant to both prongs of the duress defense.
Our conclusion is further supported by the decisions of
the vast majority of courts that have long held that expert
testimony on battered woman syndrome can be relevant in the
analogous context of self-defense. See, e.g., Humphrey, 921
P.2d at 8-9; State v. Kelly, 478 A.2d 364, 377-78 (N.J. 1984);
Ibn-Tamas v. United States, 407 A.2d 626, 631-35 (D.C.
1979); see also Janet Parrish, Trend Analysis: Expert
Testimony on Battering and Its Effects in Criminal Cases, in
DOJ REPORT pt. II, at 19, 28. The elements of self-defense
are similar to the elements of duress: To establish a claim of
self-defense in most jurisdictions, a defendant must prove that
she reasonably believed her use of force was necessary to
prevent imminent death or serious bodily harm. See 2
LAFAVE, SUBSTANTIVE CRIMINAL LAW § 10.4. Thus, if
battered woman syndrome can be relevant to prove self-
17
defense (as virtually all courts accept), it likewise should be
relevant to prove duress.5
In sum, we conclude that expert testimony on battered
woman syndrome may be admissible as a general matter to
prove duress because such testimony can be reliable and can
be relevant to both prongs of the duress defense.
2
The next question is whether expert testimony on
battered woman syndrome would have been reliable and
relevant in Nwoye’s case. We conclude that it would have
been.
To begin with, the Government does not argue that the
expert testimony from Dr. Giunta that Nwoye proffered in her
post-conviction proceeding was unreliable. And we have no
reason to question its reliability.
Nwoye’s expert testimony, moreover, would certainly
have been relevant to Nwoye’s defense. This Court suggested
as much on Nwoye’s direct appeal by noting the conspicuous
absence of expert testimony on battered woman syndrome at
Nwoye’s trial. Nwoye, 663 F.3d at 465. And the Government
does not dispute that Nwoye’s trial testimony strongly
5
The Government points out that the defenses of self-defense
and duress are distinct in one sense: Defendants who act in self-
defense often injure morally culpable individuals – the initial
aggressors – while defendants who act under duress often injure
blameless third parties. Thus, the Government argues, the duress
defense should be more narrowly circumscribed in order to protect
innocent third parties. Maybe so. But the proper response to such a
fundamental mismatch between the defenses would be to make
wholesale changes to the duress defense rather than to tweak on an
ad hoc basis the kinds of evidence that are relevant to duress.
18
suggested that she had been a victim of a battering
relationship. An expert on battered woman syndrome could
therefore have helped the jury assess the reasonableness of
Nwoye’s actions, as we described above.
3
The next question is whether expert testimony on
battered woman syndrome would have entitled Nwoye to a
duress instruction in this case. On Nwoye’s direct appeal,
when we rejected Nwoye’s claim that she was entitled to a
duress instruction, we pointed specifically to the absence of
“expert witnesses testifying to the effects of isolation,
financial dependence, or estrangement from family
members.” Id. We now conclude that the introduction of
such testimony at Nwoye’s trial would have entitled Nwoye
to a duress instruction.
Perhaps most critically, expert testimony on the
likelihood of retaliatory violence upon separation could have
provided a plausible explanation for why Nwoye failed to
extricate herself from the extortion scheme. Nwoye may have
feared that any attempt to leave Osuagwu would have resulted
in still greater violence. Moreover, Nwoye may have
reasonably believed that reporting Osuagwu to the police (or
others) would have been unlikely to result in his immediate
arrest and would have therefore placed her at greater risk in
the interim. Thus, Nwoye’s testimony concerning Osuagwu’s
abuse, supplemented by expert testimony on battered woman
syndrome, would have constituted “sufficient evidence from
which a reasonable jury could find” for Nwoye on a theory of
duress. Id. at 462.
It may be helpful here to take a step back so that the
reader does not miss the forest for the trees. The concept of
battered woman syndrome fits this case to a T. A woman was
19
beaten repeatedly by her boyfriend. Some outsiders may
question why she didn’t just leave her boyfriend. But the
expert testimony would help explain why. For the
Government to come in now and say that such expert
testimony, combined with Nwoye’s own testimony about the
beatings, still would not entitle her to a duress instruction is to
say in essence that battered woman syndrome does not matter,
at least in duress cases. We do not agree with that suggestion.
B
To this point, we have concluded that expert testimony
on battered woman syndrome would have entitled Nwoye to a
jury instruction on duress. To make out her claim of
prejudice for purposes of the ineffective-assistance-of-counsel
inquiry, Nwoye also must show that expert testimony on
battered woman syndrome, together with the duress
instruction, would have created a “reasonable probability that
. . . the result of the proceeding would have been different.”
Strickland, 466 U.S. at 694. To repeat what we noted above,
Nwoye does not need to show that the expert testimony and
the jury instruction “more likely than not” would have
produced an acquittal. Id. at 693. She must demonstrate only
“a probability sufficient to undermine confidence” in the
verdict. Id. at 694.
It can be difficult for a reviewing court (whether a district
court or a court of appeals) to determine how additional
evidence or an additional jury instruction would have affected
a trial. It is inherently a speculative exercise. In this case,
however, we conclude that Nwoye has demonstrated a
reasonable probability that the jury would have had a
reasonable doubt respecting her guilt.
At trial, Nwoye admitted to participating in the extortion
scheme, but said she did so at the direction of Osuagwu.
20
Based on that testimony, Nwoye’s only possible defense was
that she acted under duress. But due in part to trial counsel’s
failure to present expert testimony on battered woman
syndrome, Nwoye was denied a jury instruction on duress.
And the District Court specifically precluded Nwoye’s
counsel from mentioning duress during closing arguments.
Thus, the failure of Nwoye’s counsel to present expert
testimony on battered woman syndrome deprived Nwoye of
any viable legal avenue to acquittal.
With an instruction on duress, together with expert
testimony on battered woman syndrome, there is at least a
reasonable probability that the jury “would have had a
reasonable doubt respecting guilt.” Id. at 695. Under very
similar circumstances, the Sixth Circuit held that an attorney’s
failure to introduce expert testimony on battered woman
syndrome prejudiced a defendant who pled guilty to robbery
and related charges. See Dando v. Yukins, 461 F.3d 791 (6th
Cir. 2006). The Court reasoned that, if counsel had
introduced such testimony, the defendant would have had a
sufficient evidentiary basis for a duress defense. And if the
defendant had proceeded to trial on a theory of duress, there
would have been a “likelihood of a favorable outcome at trial”
such that the defendant’s counsel would not have
recommended that the defendant take a guilty plea. Id. at
800-02.
A similar analysis applies here. The duress instruction
would have given jurors a legal basis upon which to vote not
guilty. And as discussed above, the expert testimony on
battered woman syndrome would have supported both
elements of Nwoye’s duress defense: the imminence of the
threat and the absence of reasonable alternatives. In addition,
expert testimony on battered woman syndrome would have
bolstered the credibility of Nwoye’s testimony about
21
Osuagwu’s abuse. Jurors faced with testimony from a
battered woman concerning her abuse and its effects may
doubt the testimony because they do not believe that a woman
subject to such abuse would stay with her abuser without
alerting police or others. Expert testimony on battered
woman syndrome could have helped Nwoye “dispel the
ordinary lay person’s perception that a woman in a battering
relationship is free to leave at any time.” Humphrey, 921 P.2d
at 9; see also Kelly, 478 A.2d at 377.
Those factors add up to a reasonable probability that the
jury would have had a reasonable doubt respecting guilt if
expert testimony on battered woman syndrome had been
presented at Nwoye’s trial. The jury of course could still have
convicted Nwoye; for example, the jury could have
disbelieved that Nwoye was telling the truth about the abuse
in the first place or could have been unpersuaded by the
expert testimony. But for present purposes on appeal, we
have no basis to question that Nwoye has told the truth about
being abused by Osuagwu or to question the expert testimony.
(The Government does not argue otherwise.) On this record,
we conclude that Nwoye’s testimony plus the expert
testimony plus the duress instruction create at least a
reasonable probability that the jury would have had a
reasonable doubt respecting guilt. Nwoye was prejudiced by
her counsel’s failure to present expert testimony on battered
woman syndrome.
***
The District Court considered only the prejudice element
of the inquiry into ineffective assistance of counsel. The
District Court found no prejudice. Because we have
concluded that Nwoye was prejudiced by the failure of her
trial counsel to introduce expert testimony on battered woman
22
syndrome, we remand for the District Court to consider in the
first instance the other prong of the ineffective-assistance-of-
counsel standard: whether the performance of Nwoye’s
counsel fell below an objective standard of reasonableness. If
counsel’s performance was constitutionally deficient, then
Nwoye will have established ineffective assistance of counsel
and will be entitled to have her conviction vacated. We
reverse the judgment of the District Court and remand for
further proceedings.
So ordered.
SENTELLE, Senior Circuit Judge, dissenting: This case
returns to us on a petition for writ of coram nobis, testing a
criminal conviction previously affirmed by us on direct appeal
in United States v. Nwoye, 663 F.3d 460 (D.C. Cir. 2011). In the
original version, we considered an argument that the trial court
had erred in refusing a defense request for an instruction on a
duress defense. We affirmed the district court, holding that “a
defendant is only entitled to an instruction on a theory of duress
if there is ‘sufficient evidence from which a reasonable jury
could find’ for the defendant on that theory.” Id. at 462 (quoting
United States v. Akhigbe, 642 F.3d 1078, 1083 (D.C. Cir.
2011)). Appellant Nwoye returns, alleging that her trial counsel
rendered ineffective assistance in not presenting expert
testimony on the battered woman syndrome and that she was
prejudiced thereby. The district court was unconvinced, for
reasons expressed in a most able opinion, United States v.
Nwoye, 60 F. Supp. 3d 225 (D.D.C. 2014). My colleagues on
this court reason differently than the district judge. I consider
the district judge’s reasoning the more compelling, and I will
quote from it extensively herein.
As the majority sets forth, to prevail on the claim for
ineffective assistance of counsel, a petitioning defendant must
establish two elements: “(i) ‘counsel’s representation fell below
an objective standard of reasonableness’ and (ii) ‘the deficient
performance prejudiced the defense.’” Maj. Op. at 9 (quoting
Strickland v. Washington, 466 U.S. 668, 687-88 (1984)). The
district court pretermitted the first question and proceeded to
determine that counsel’s decision not to introduce expert
evidence on the subject of battered woman syndrome was not
prejudicial. Because the majority deems the district court to
have erred in the second part of its decision, we finish in the odd
posture of sending the matter back for the resolution of the
preliminary question. Be that as it may, in my view, the district
2
court did not err on the second prong, so like the majority, I will
find it unnecessary to draw a resolution on the first.
I would note in passing that in judging the degree, if not the
kind, of counsel’s ineffectiveness, we have never before this
case ruled on the admissibility of expert testimony on battered
woman syndrome to a claimed defense of duress. As the district
court noted, “Traditionally, expert testimony on BWS has been
limited to cases where a defendant puts forward the affirmative
defense of self-defense.” 60 F. Supp. 3d at 237. While I do not
dispute the majority’s extension of the relevance to a duress
defense in appropriate cases, I do find it less than shocking that
the trial counsel did not more vigorously pursue the possibility
in a pioneering posture than would have been the case had the
defense been self defense. Be that as it may, as the majority
notes, the issue directly confronting us is whether counsel’s
failure, construed as ineffective assistance, “prejudiced the
defense.” As the majority states, “Nwoye . . . must show that
expert testimony on battered woman syndrome, together with
the duress instruction, would have created a ‘reasonable
probability that . . . the result of the proceeding would have been
different.’” Maj. Op. at 19 (quoting Strickland, 466 U.S. at 694).
Like the district court, I am not convinced that she has made this
showing.
To be entitled to an instruction on the defense of duress, a
defendant must introduce at least some evidence on the two
elements of the defense: (1) that the defendant acted under the
threat of immediate death or serious bodily injury, and (2) that
the defendant had no reasonable legal alternative to committing
the crime. See United States v. Gaviria, 116 F.3d 1498, 1531
(D.C. Cir. 1997). Importantly, in the context of the present case,
defendant must not only show some evidence on these two
elements, she must also establish that if counsel had proffered
the evidence (which he did except for the expert testimony), the
3
judge had allowed it in, and the judge had based instruction
thereon, the jury result would have been different. The record
simply does not support the majority’s conclusion on this cluster
of issues.
As the district court put it, the trial judge
provided the defendant with ample opportunity to present
evidence in support of both of these necessary elements.
She permitted the defendant to testify at length about the
abuse that she said she had suffered at the hands of Mr.
Osuagwu. In addition to Ms. Nwoye’s testimony about her
abuse, the judge also heard evidence that undercut
defendant’s theory of the case, including evidence that (1)
Mr. Osuagwu frequently left the D.C.-Maryland area
without her, taking trips to California that lasted days or
weeks; (2) Ms. Nwoye left Mr. Osuagwu and returned to
her husband in the summer of 2006 without incident; and
(3) Ms. Nwoye eventually contacted the Nigerian security
services regarding Mr. Osuagwu’s criminal behavior.
60 F. Supp. 3d at 240.
Based on this evidence, the district court did not submit the
duress instruction as requested. The jury returned a verdict of
guilty. Then, as the district court noted, this court reviewed and
affirmed the district court’s decision.
In so doing, we
noted that Ms. Nwoye regularly left her home to attend
nursing school classes and to work at the hospital and was
thus “physically separated” from Osuagwu. [United States
v. Nwoye, 663 F.3d at 463.] She also met alone with Dr.
Iweala and did not tell him that she was being forced to
4
extort money from him. Id. Most importantly, the court of
appeals emphasized that “Osuagwu spent nearly two weeks
in California, thousands of miles away from Nwoye,”
giving her more than enough opportunity to notify law
enforcement. Id. at 463-64. “[A] defendant with such
‘countless opportunities to contact law enforcement
authorities or [to] escape the perceived threats’ cannot as a
matter of law avail herself of the duress defense.” Id. at
464 (quoting United States v. Scott, 901 F.2d 871, 874 (10th
Cir. 1990)).
Id. at 231. Further, with direct reference to Nwoye’s claim that
she may have suffered from BWS, we explained that
Nwoye suggests the mere whiff of [BWS] arising from
these facts should alter the duress determination. . . . [But]
her theory is devoid of the other usual indicia supporting a
BWS defense—expert witnesses testifying to the effects of
isolation, financial dependence, or estrangement from
family members. Indeed, as discussed earlier, Nwoye had
many alternative sources of protections and support[,] . . .
[including] access to relatives, classmates, and teachers
with whom she could seek refuge. [Furthermore,] [s]he was
not under constant visual surveillance. The conspiracy in
which she participated lasted for months, . . . [and there
were] weeks in which Osuagwu was thousands of miles
away.
United States v. Nwoye, 663 F.3d at 465.
So seeing, we affirmed. Years later came the present
petition for writ of coram nobis.
The petition originally came back to the trial judge. Due to
her having suffered serious injury, the matter was reassigned to
5
a different judge whose decision we now review. The district
court heard the testimony of four witnesses, including a forensic
psychologist, the defendant herself, trial counsel, and the public
defender now representing Ms. Nwoye.
With the trial record and the testimony of the four witnesses
before him, the district judge applied the two-step analysis from
Strickland, 466 U.S. at 687-88, and correctly noted that “[t]he
Strickland prejudice prong requires only a reasonable
probability—that is, ‘a probability sufficient to undermine
confidence in the outcome.’” 60 F. Supp. 3d at 234 (quoting
Strickland, 466 U.S. at 694).
The district judge then explained that “‘the [duress]
standard is . . . partially objective; the defense is not established
simply by the fact that the defendant was coerced; he must have
been coerced in circumstances under which a person of
reasonable firmness in his situation would likewise have been
unable to resist.’” Id. at 238 n.6 (quoting Model Penal Code
§ 2.09, Explanatory Note (2001)) (emphasis added by the
district court).
Applying these standards, the district court reasoned:
In the context of duress, this means that a judge must
instruct a jury on the defense of duress only when the
defendant presents at least some evidence on both of the
necessary elements of the defense. “To prevail on a duress
defense, a defendant must convince the jury that (1) she
acted under the threat of immediate death or serious bodily
injury, and (2) that she had no reasonable legal alternative
to committing the crime, i.e., no chance both to refuse to do
the criminal act and also to avoid the threatened harm.”
Id. at 239 (quoting United States v. Nwoye, 663 F.3d at 467
6
(Tatel, J., dissenting) (other internal quotation marks omitted)).
With that and further reasoning, the court concluded, in my
view reasonably, that “[b]ecause no amount of expert testimony
could rectify [the] evidentiary deficiency, . . . there was no
prejudice to the defendant by virtue of her lawyer’s failure to
offer expert testimony on BWS for presentation to” the trial
court. Id. at 241. “Rather, in order for the defendant to be
entitled to a duress instruction, she must adduce evidence at trial
showing that she herself in fact had no reasonable legal
alternative.” Id. (citations omitted).
As did the district court, I recognize, as the majority points
out, that this court in the direct appeal noted the lack of expert
testimony on battered woman syndrome. Maj. Op. at 7, 11.
However, I am not sure that I share my colleagues’ conviction
that the prior opinion “stressed” that omission, but rather listed
it among a series of facts. Id. at 7. I finally suggest that we
should remember what we are reviewing. The determination of
probabilities of the effect of counterfactual circumstances
partakes more of a decision of fact than of law. It is hornbook
learning that we review factual decisions of district courts with
deference. See generally Anderson v. City of Bessemer City,
NC, 470 U.S. 564 (1985). Indeed, Rule 52(a)(6) of the Federal
Rules of Civil Procedure provides that “[f]indings of fact . . .
must not be set aside unless clearly erroneous, and the reviewing
court must give due regard to the trial court’s opportunity to
judge the witnesses’ credibility.” While we may not be literally
reviewing “a finding,” we are essentially reviewing a factual
determination. In this case two district judges have had the
benefit of the credibility of witnesses and the experience of
making factual determinations. I would afford them that same
deference.
I therefore respectfully dissent.