IN THE COURT OF CRIMINAL APPEALS
OF TEXAS
NO. PD-0635-19
JAMES RAY HAGGARD, Appellant
v.
THE STATE OF TEXAS
ON APPELLANT’S PETITION FOR DISCRETIONARY REVIEW
FROM THE NINTH COURT OF APPEALS
LIBERTY COUNTY
HERVEY, J., delivered the opinion of the Court in which KEASLER,
RICHARDSON, NEWELL, and WALKER, JJ., joined. YEARY, J., filed a concurring
opinion. SLAUGHTER, J., filed a dissenting opinion. KELLER, P.J., and KEEL, J.,
concurred.
OPINION
The Sixth Amendment to the United States Constitution guarantees a criminal
defendant the right to confront adverse witnesses. The issue here is whether the
Confrontation Clause was violated when the trial judge allowed Suzanne DeVore, a
Sexual Assault Nurse Examiner (SANE), to testify against James Ray Haggard
Haggard–2
(Appellant) from Montana using a two-way video system.
Because we conclude that admitting DeVore’s remote testimony violated the
Confrontation Clause, we will reverse the judgment of the court of appeals holding
otherwise and remand the cause to the court of appeals to conduct a new harm analysis.
FACTS
On October 5, 2013, the victim’s mother, Traci, took her two children, M.W. and
A.W., to Haggard’s house to play with his stepson and daughter and spend the night,
something that happened often. Traci was life-long friends with Haggard, and her
daughters referred to him as “uncle.” At the time of the alleged assault, M.W. (the victim)
was fifteen years old and A.W. was ten years old. M.W. testified that there were only two
bedrooms in the house and that A.W. slept in the living room with Haggard’s daughter,
that Haggard’s stepson slept in the second bedroom, and that she slept in Haggard’s room.
M.W. told the DeVore that Haggard “pulled [her] back to his bedroom.”1 But at trial, she
1
M.W. told DeVore in a dictated statement (which we reproduce here without alteration)
that,
He’s my uncle & I didn’t think he’d do anything perverted. He told me to take my
shirt off. I told him I felt uncomfortable. He wasn’t forcing me but he took my
bottoms off & started playing with my boobs with his fingers & started licking
them. On my right boob, I have a hickey. He went down on me, his mouth inside
& outside my vagina, & then started doing the ‘dirty deed’, his penis in my
vagina. He heard someone walking down the hall & said, ‘Hurry up & put on your
clothes’. He kept saying I better not tell because he’d lose his baby & if anyone
found out he’d go to jail, & it would tear apart the whole family. “He said, ‘You’d
better not tell your mom’. I had an hour of sleep. I woke up, called my mom, &
Haggard–3
said that she slept in Haggard’s bed because she did not want to sleep on the floor and did
not think that Haggard would do anything “perverted.”
M.W. told DeVore that, shortly after they got in bed,2 Haggard asked her to take
off her clothes,3 which she did, then he sexually assaulted her. M.W. said that Haggard
digitally manipulated her breasts and put his mouth on them, that he put his mouth on her
sexual organ, and that he penetrated her vagina with his penis. The assault ended, M.W.
told DeVore, when Haggard told her to get dressed and leave because he thought he heard
someone walking down the hallway towards his room. M.W. also told DeVore that
Haggard “cleaned himself off with my shirt.” M.W. testified that, after she left Haggard’s
room, she made Haggard’s stepson sleep in the living room with the girls and that she
slept in the second bedroom bed by herself. She also testified that, after the assault, she
texted her friend Brittany and called her friend Austin.
M.W. waited until the following morning to call her mother to pick her and her
sister up. When she called, she was crying and upset, but did not tell Traci what
told her to come get me. She asked if he tried to touch me & I said yes he did. My
mom asked me because my sister who is 6 doesn’t want to go over there anymore
& my mom thinks my uncle’s boy who is 10 touched her.”
Later in the report, DeVore noted that M.W. also told her that “[h]e cleaned himself off with my
shirt.”
2
M.W. later testified that Haggard showed her pornography before he asked her to take
off her clothes. She did not mention that to DeVore.
3
M.W. was wearing black shorts, a gray shirt, and two bras—a “regular bra” and a sports
bra. She told DeVore that she was wearing two bras because she had been playing softball that
day.
Haggard–4
happened. Traci testified that M.W. sounded distressed. M.W. called Traci multiple times,
asking her to “hurry up,” so Traci called Linda Brackin-Barton (Barton), M.W.’s aunt, to
pick the kids up since she lived closer to Haggard.4 When Barton and the kids returned to
Traci’s house, Barton said that M.W. told her and Traci what happened. Barton said that
M.W. went through the story several times with her and Traci and that “[M.W.] was just
talking. She seemed to want to talk about what had happened and I think we were just
kind of letting her talk . . . .” M.W. testified that Traci was in the room the entire time, but
she also said that she never told her mother the complete story. At trial, Traci testified that
she was only in and out of the room and only heard “bits and pieces.”
No one called the police, but Traci put M.W.’s clothes in a Ziploc bag because she
saw “stuff on them.” Later that day, Traci took M.W. to a friend’s birthday party. The
next day, Traci went to work and sent M.W. to school even though she did not want to go.
M.W. testified that she was going to call her father and ask him to pick her up from
school because she did not feel well, but Barton unexpectedly showed up and took her to
the hospital for a SANE examination. Barton testified that she read that victims of sexual
assaults should be physically examined. Barton said that she called Traci on the way to
the school and that Traci did not go to the hospital until later in the day because she was
at work.
DeVore performed the SANE examination and wrote down M.W.’s account of
4
Barton testified that Traci asked her to pick up the children because Traci was too upset.
Haggard–5
what happened.5 According to DeVore’s notes, M.W. was calm and cooperative and made
good eye-contact. M.W. told DeVore that Haggard penetrated her vagina, but not her
mouth or anus. She also said that she was not sure if Haggard ejaculated on her. DeVore
physically examined M.W. and found no trauma to M.W.’s labia majora, labia minora,
hymen, vagina, cervix, perineum, or anus. She did, however, find a red and blue bruise
approximately 1.5 centimeters by 1 centimeter on M.W.’s right breast.
Later that evening, M.W. was examined by a forensic examiner at Bridgehaven
Child’s Advocacy Center. M.W. told the examiner that Haggard asked her to play with
his penis before they had sex and that, after a while, “semen went everywhere.” She also
said that she “felt wetness on my hand and on my face. So, I wiped it off with my shirt.”
Sergeant Stephen Clappart, who was present for the interview, noted that M.W. was
“jovial” when he spoke to her. He also noted that M.W. gave the examiner a slightly
different account of events.
Jessica Lake, a serologist at the Department of Public Safety crime laboratory in
Houston, examined the contents of the SANE kit. The kit contained vaginal swabs, labia
majora swabs, labia minor swabs, anal swabs, swabs taken from M.W.’s breasts, head-
hair combings, a pair of underwear, a bra, a sports bra, and a shirt. Lake did not find the
semen on the vaginal swabs, labia swabs, or anal swabs. Lake found areas of interest on
the underwear and shirt to test, but both tests were negative for the presence of semen.
5
See supra, note 1.
Haggard–6
Andrea Smith, Lake’s supervisor, was DNA forensic examiner. She issued her first
report in September 2014. Smith found a single DNA profile consistent with M.W. on the
vaginal, labia majora, labia minor, and anal swabs. Testing of the left-breast swabs was
inconclusive, but a partial DNA profile was extracted from the right-breast swabs using a
MiniFiler™ amplification kit.6 The partial profile was consistent with a two-person
mixture, from which neither M.W. nor Haggard could be excluded as contributors. Smith
said that the probability of selecting an unrelated person at random who could have
contributed to the partial profile was 1 in 6 for Caucasians, 1 in 5 for Blacks, and 1 in 4
for Hispanics. The approximate world population at the time of testing was 7 billion
people.
In 2017, Smith reinterpreted the DNA data using new testing guidelines and
software.7 Her results were significantly more inculpatory. She found that it was 339
billion times more likely that M.W. and Haggard contributed to the mixed DNA on the
right-breast swab than M.W. and some other unknown and unrelated individual.8 Smith
6
When genetic samples are degraded or low-level, scientists sometimes use a MiniFiler™
chemistry kit to amplify the genetic material, which might yield a partial DNA profile where
traditional methods fail.
7
In the past, DPS technicians issued a single report even when different chemistry kits
were used (e.g., a traditional kit and a MiniFiler™ kit) so long as both kits were used to analyze
the same area of interest (e.g., a stain on a shirt). The updated guidelines require technicians to
issue a separate report for each kit used.
8
Before DPS began using the new software, the results of DNA analyses were expressed
in the form of a “combined probability of inclusion” statistic (e.g., 1 in 4 Caucasians). The new
software, however, issues results in terms of the “likelihood ratio” (LR) that a particular person
contributed to a particular DNA mixture.
Haggard–7
was able to develop a partial DNA profile consistent with a two-person mixture from the
left-breast swab data, but testing was inconclusive as to whether Haggard was a
contributor. Smith was also able to develop a partial DNA profile from the right-breast
swab data (without using a MiniFiler™ kit), which she was previously unable to do, and
concluded that it is 219 quadrillion times more likely that M.W. and Haggard contributed
to the mixture than M.W. and some other unrelated and unknown individual.9
CONFRONTATION CLAUSE
a. Background
By the time of trial, DeVore had moved to Montana, but she told the State that she
would voluntarily travel to Texas to appear in court and testify. The State agreed to
reimburse her expenses, but did not subpoena her. The Friday afternoon before trial
began, however, DeVore notified prosecutors that she would not voluntarily appear and
testify.
The following Monday, the State asked the trial court to allow DeVore to testify
from Montana via FaceTime. According to the State, DeVore’s testimony was essential
because only she could prove-up the chain of custody of the SANE kit and its contents. If
she did not testify, the State alleged, the DNA reports would be inadmissible. The trial
court granted the State’s motion, and DeVore testified the next day after the court
9
One reason for the starkly different results is that the MiniFiler™ kit relies on only nine
loci to develop a DNA profile, but a traditional chemistry kit relies on 16 loci, which is almost
60% more genetic information.
Haggard–8
overruled another defense objection.
DeVore testified that even though she initially agreed to appear in court and that
reimbursement of her expenses would be sufficient for her to voluntarily appear and
testify, she changed her mind at the last minute for economic and personal reasons. She
said that, although the State would reimburse her expenses, it was not going to pay her to
testify and neither would her employer. (DeVore testified that she was a self-employed
consultant.) DeVore, however, never said that she was incapable of traveling to Texas due
to financial constraints, or even if she could, traveling to Texas would cause an undue
financial burden. DeVore testified that it was inconvenient timing for her because she
flew to Houston the week before to testify at a different trial, and she had to fly to
Houston again the weekend after the instant trial to be with an ill family member.
The State brought up that it did not subpoena DeVore and that DeVore’s husband
was a lawyer and told her that she did not have to voluntarily appear and testify unless she
was subpoenaed. But the State focused on how it had insufficient time to subpoena
DeVore due to her late notification, even though it had ample time to subpoena her before
trial but never did. We also note that the State never sought a continuance before asking
the trial court to allow DeVore to testify remotely.
b. Applicable Law
The Sixth Amendment Confrontation Clause provides that “[i]n all criminal
prosecutions, the accused shall enjoy the right . . . to be confronted with the witnesses
Haggard–9
against him[.]” U.S. CONST. amend. VI. The Supreme Court long held that the
Confrontation Clause protected a criminal defendant’s right to physically confront those
who testify against him. Coy v. Iowa, 487 U.S. 1012, 1017 (1988) (quoting Pennsylvania
v. Ritchie, 480 U.S. 39, 51 (1987) (plurality op.)).
For example, in Coy, the issue was whether allowing two child victims of sexual
abuse to testify with a screen placed between the victims and Coy violated the
Confrontation Clause. Id. at 1014. The victims could not see Coy, but he could hear them
and see their faint outlines.10 Id. at 1014–15. The State argued that “the confrontation
interest at stake . . . was outweighed by the necessity of protecting victims of sexual
abuse,” but the Supreme Court held that the arrangement violated the Confrontation
Clause. Id. at 1020. The majority agreed that physical, face-to-face confrontation might
be properly dispensed with if doing so furthered an important public policy. Id. But it
explained that it did not need to decide the issue because, even if such an exception
existed, there were no individualized findings that the witnesses “needed special
10
An Iowa statute provided for child-witness testimony via closed-circuit television or
behind a screen. That statute read in relevant part,
The court may require a party be confined to an adjacent room or behind a screen
or mirror that permits the party to see and hear the child during the child’s
testimony, but does not allow the child to see or hear the party. However, if a
party is so confined, the court shall take measures to insure (sic) that the party and
counsel can confer during the testimony and shall inform the child that the party
can see and hear the child during testimony.
IOWA CODE § 910A.14 (1987).
Haggard–10
protection . . . .” Id. at 1021. Two years later, the Supreme Court was squarely presented
with the exception question.
In Maryland v. Craig, 497 U.S. 836 (1990), the trial court allowed a child victim
of sexual abuse to testify via a one-way, closed-circuit television system pursuant to a
Maryland statute. Unlike the Iowa statute in Coy, the Maryland statute required the trial
court to make a finding that the “‘testimony by the child victim in the courtroom will
result in the child suffering serious emotional distress such that the child cannot
reasonably communicate.’” Id. at 840 n.1. The Supreme Court held that face-to-face
confrontation was properly dispensed with because (1) it was to further the “important
public policy” of protecting the “physical and psychological well-being of child abuse
victims,” which outweighed Craig’s right to face his accuser in court, and (2) “the
reliability of the testimony [was] otherwise assured” because the child victim testified
under oath, was fully cross-examined, and the judge, jury, and Craig could see the victim
while she testified. Id. at 850.
Craig, however, was not without limits. The court emphasized that a trial court
must make a “finding of necessity” that is “case-specific” and that the trial court must
hear evidence,
[I]f the State makes an adequate showing of necessity, the state interest in
protecting child witnesses from the trauma of testifying in a child abuse
case is sufficiently important to justify the use of a special procedure that
permits a child witness in such cases to testify at trial against a defendant in
the absence of face-to-face confrontation with the defendant.
Haggard–11
The requisite finding of necessity must of course be a case-specific one:
The trial court must hear evidence and determine whether use of the
one-way closed circuit television procedure is necessary to protect the
welfare of the particular child witness who seeks to testify.
Id. at 855. In recognizing the need for a necessity finding, as the above quote shows, the
Court acknowledged that the physical and psychological well-being of the child might not
always be sufficient to outweigh a defendant’s right to physical, face-to-face
confrontation. Id. at 852.
Since Craig was decided, this Court has required a necessity finding in every case
in which we have considered a Confrontation Clause challenge to the cross-examination
of a witness via two-way video system. Marx v. State, 987 S.W.2d 577, 580–81 (Tex. Crim.
App. 1999) (applying Craig to the use of a two-way video system and requiring a necessity
finding); Gonzales v. State, 818 S.W.2d 756, 764 (Tex. Crim. App. 1991) (same). And we are
not alone in doing so. Many federal circuit courts of appeals and state supreme courts
have reached the same conclusion.11 See, e.g., United States v. Cotto-Flores, 970 F.3d 17,
25 (1st Cir. 2020) (applying Craig to two-way video and requiring a necessity finding);
United States v. Carter, 907 F.3d 1199, 1206 (9th Cir. 2018) (same); United States v. Abu
11
In an unpublished opinion, the Eleventh Circuit held that it was a reasonable application
of federal law to not apply Craig to two-way video. Fuster-Escalona v. Fla. Dep’t of Corr., 170
F. App’x 627, 629–30 (11th Cir. 2006). Even if we were to regard this unpublished opinion as
authority, however, it is clearly an outlier. Later that same year, the Eleventh Circuit applied
Craig in a published opinion. Yates, 438 F.3d at 1313. And while we acknowledge that the
Second Circuit Court of Appeals uses different language, it still requires a necessity finding
albeit one that demonstrates “exceptional circumstances.” United States v. Gigante, 166 F.3d 75,
81 (2d Cir. 1999).
Haggard–12
Ali, 528 F.3d 210, 240 (4th Cir. 2008) (same); United States v. Yates, 438 F.3d 1307,
1313 (11th Cir. 2006) (same); United States v. Bordeaux, 400 F.3d 548, 554 (8th Cir.
2005) (same); Lipsitz v. State, 442 P.3d 138, 140 (Nev. 2019) (same); State v. Thomas,
2016-NMSC-024, ¶ 28, 376 P.3d 184, 194 (same); State v. Rogerson, 855 N.W.2d 495,
504 (Iowa 2014) (same); State v. Stock, 2011 MT 131, ¶¶ 25, 30, 256 P.3d 899, 904
(same); Bush v. State, 2008 WY 108, ¶ 52, 193 P.3d 203, 215 (same); Harrell v. State,
709 So. 2d 1364, 1369 (Fla. 1998) (same).
Forced to concede that this Court has applied the Craig test to two-way video, the
dissent contends that those cases are outdated and that we should “focus on the realities of
the world we live in today . . . .” Dissenting Op. at 10 n.11. Given that Gonzales, Marx,
and this case all utilize live two-way video, apparently, the dissent means that the reality
of the world we live in today is that there are more televisions and their screens are
bigger. Id. at 9 (stating that the witness’s image was projected onto video screens at both
counsel’s tables, on the judge’s video screen, and on a 60-inch television for the jury); see
id. at 8 (stating that Haggard “made no objections to the positioning of the camera or to
the number and size of screens utilized.”). We are not aware of a reported decision, and
the dissent does not cite one, in which the number of televisions projecting the witness’s
visage or the size of the televisions control a Confrontation Clause inquiry. Writing about
two-way video and the Confrontation Clause, however, we note that Justice Scalia said,
I cannot comprehend how one-way transmission (which Craig says does not
ordinarily satisfy confrontation requirements) becomes transformed into
Haggard–13
full-fledged confrontation when reciprocal transmission is added. As we
made clear in Craig, . . . a purpose of the Confrontation Clause is ordinarily
to compel accusers to make their accusations in the defendant’s
presence—which is not equivalent to making them in a room that contains a
television set beaming electrons that portray the defendant’s image. Virtual
confrontation might be sufficient to protect virtual constitutional rights; I
doubt whether it is sufficient to protect real ones.
Order of the Supreme Court, 207 F.R.D. 89, 91 (2002) (statement of Scalia, J.) (emphasis
added).
Later, the Supreme Court decided Crawford v. Washington, 541 U.S. 36, 61–65
(2004), in which it abandoned its previous reliability-based approach (the second prong of
the Craig test) and “adopted a fundamentally new interpretation of the confrontation
right . . . .” It said that the Confrontation Clause is “a procedural rather than a substantive
guarantee [that] commands, not that evidence be reliable, but that reliability be assessed
in a particular manner: by testing in the crucible of cross-examination.” Id. at 61.
Although Crawford was about the admission of out-of-court statements, not the
dispensation of face-to-face confrontation, some have argued that Crawford implicitly
overruled Craig, or at least the reliability prong.12 See Coronado v. State, 351 S.W.3d
315, 321 (Tex. Crim. App. 2011) (discussing how post-Crawford cases have changed the
Confrontation Clause landscape).
12
See, e.g., Horn v. Quarterman, 508 F.3d 306, 318–19 (5th Cir. 2007); State v. Arroyo,
935 A.2d 975, 992 n. 18 (Conn. 2007); State v. Henriod, 2006 UT 11, ¶ 17, 131 P.3d 232,
237–38; Stock, 256 P.3d at 904 (stating that no court has “concluded Crawford overruled
Craig.”); Roadcap v. Commonwealth, 653 S.E.2d 620, 625 (Va. 2007) (“As nearly all courts and
commentators have agreed, Crawford did not overrule Craig.”).
Haggard–14
c. Analysis
There is no doubt that the role of reliability in a Confrontation Clause analysis has
been called into question. But we need not resolve that issue here because our holding is
not based on the reliability of DeVore’s testimony, but rather whether allowing her to
testify remotely “furthered an important public policy.”13 We conclude that it did not.
First, however, we must address the lack of a case-specific finding and the
requirement that the court hear evidence. There was a discussion between the judge and
attorneys and arguments were made, but the judge heard no evidence and made no case-
specific finding. The closest thing to a necessity finding in the record is when the judge
suggested that DeVore should be allowed to testify because the State did not have time to
subpoena her given her late notice. The judge told defense counsel that Craig was
inapplicable and that it was allowing DeVore to testify remotely because she was only
testifying as an expert and because the procedures ensured the reliability of her
testimony.14 But the fact that DeVore was testifying “only” as an expert does not
13
See Gonzales, 818 S.W.2d at 764 (requiring a necessity finding before authorizing the
dispensation of physical, face-to-face confrontation in favor of a two-way video system).
14
The dissent appears to argue that virtual confrontation is superior to physical
confrontation. Dissenting Op. at 10 (“[I]t is hard to argue against the fact that current technology
allows for the opportunity to see and hear a witness better through two-way video than if that
witness was testifying in the courtroom.”) (emphasis added). But the facts of this case belie the
dissent’s assertion. There were numerous technical issues with DeVore’s testimony. At one
point, DeVore could not hear questioning from the prosecutor, and a short time later the signal
was interrupted and no one in the courtroom could see DeVore. Later, defense counsel objected
and reminded the trial judge that the judge said he would revisit his ruling if there were any
“hiccups.” The trial court overruled that objection. There were more problems later, but the trial
court never revisited its ruling.
Haggard–15
demonstrate a necessity for the State to procure her testimony remotely.15 Further, we do
not think it is an important public policy to allow the State to procure a witness’s
testimony remotely when the State had sufficient time and ability to subpoena the witness,
and the witness was available to appear and testify, but the State chose not to.16 Not only
is that scenario far afield from Craig, such an exception could be easily abused. Craig,
497 U.S. at 852–53, 855–58.
The next question is whether the reasons given by DeVore were sufficient to
dispense with physical, face-to-face confrontation. We conclude that they are not.
Traveling to appear in court and testify can be frustrating and difficult for various
reasons: finances, hectic schedules, health issues, sheer distance, etc. But the right to
We also note that the dissent appears to take contrary positions. On the one hand, it seems
to claim that two-video is better than physical confrontation, but on the other hand, the author
also says that she and defendants should prefer physical confrontation. Id. at 10–11. Why a
defendant would prefer a physical confrontation when two-way video is supposedly better is not
clear.
15
Neither the judge nor the State cited any authority for the proposition, but our research
has revealed one case, Kansas City v. McCoy, 525 S.W.2d 336 (Mo. 1975), in which a court held
that allowing an expert witness to testify via closed- circuit television did not violate the
Confrontation Clause. Id. at 339. But the court emphasized that the case was about a municipal-
ordinance violation, which was civil in nature, and that the procedures ensured the reliability of
the testimony. This case is distinguishable because it involves a felony, not a municipal-
ordinance violation, and it was decided before Craig, so the Court did not have the opportunity to
address the necessity-finding requirement.
16
We understand that parties sometimes do not subpoena “friendly” witnesses as a
courtesy, but that practice comes with drawbacks, as this case illustrates. The dissent would
adopt the extraordinary rule that, if a defendant wants to confront adverse witnesses in person, he
must subpoena all of the State’s witnesses. We are unaware of any courts in this jurisdiction or
any other that follows such a rule, and we observe that the dissent cites no authority to support its
proposition.
Haggard–16
physical, face-to-face confrontation lies at the core of the Confrontation Clause, and it
cannot be so readily dispensed with based on the mere inconvenience to a witness.
The State cites numerous cases that it argues justify allowing DeVore to testify
remotely, but those cases deal with child victims and child witnesses, witnesses who are
too sick to travel and appear in court, witnesses who are overseas on active duty, or
witnesses who are outside the subpoena power of the State.17
HARM
Having decided that the trial court committed constitutional error when it allowed
DeVore to testify remotely, the next question is whether the court of appeals properly
addressed harm and whether it was correct that any error was harmless.
a. The Law
17
See, e.g., Marx, 987 S.W.2d at 577 (upholding trial court’s ruling permitting a child
sexual-assault victim and child witness to testify via a two-way video system because doing so
furthered the important public policy of protecting children who testify in sexual-abuse cases
from significant emotional trauma caused by the defendant’s presence); Gonzales, 818 S.W.2d at
756 (permitting a 10-year-old murder witness to testify via a two-way closed-circuit system);
Bush, 193 P.3d at 203 (permitting a witness who was hospitalized with congestive heart failure
one week before trial to testify remotely from Colorado); Harrell, 709 So. 2d at 1364 (permitting
two adult victims from Argentina to testify remotely using a two-way video system because they
lived in Argentina, beyond the subpoena power of the Court, and because the husband had health
problems that prevented him from traveling to the United States); Stevens v. State, 234 S.W.3d
748, 781 (Tex. App.—Fort Worth Aug. 23, 2007, no pet.) (permitting a 75-year-old witness who
had been hospitalized several times for “decompensated congestive heart failure, gastrointestinal
bleeding, atrial fibrillation, and vascular disease” in the year before the trial to testify remotely
from Colorado); Rivera v. State, 381 S.W.3d 710, 711 (Tex. App.—Beaumont 2012, pet. ref’d)
(permitting crime-scene investigator to testify remotely because he was on active duty in Iraq at
the time of trial); Lara v. State, No. 05-17-00467-CR, 2018 WL 3434547, at *4 (Tex.
App.—Dallas July 17, 2018, pet. ref’d) (mem. op, not designated for publication) (permitting a
witness to testify remotely because he had a heart attack the night before trial and was in the
hospital).
Haggard–17
A denial of physical, face-to-face confrontation is reviewed for harmless error.
Coy, 487 U.S. at 1021; see Chapman v. California, 386 U.S. 18, 23 (1967). Constitutional
error is harmful unless a reviewing court determines beyond a reasonable doubt that the
error did not contribute to the conviction. TEX. R. APP. P. 44.2(a). The State has the
burden, as beneficiary of the error, to prove that the error is harmless beyond a reasonable
doubt. See Deck v. Missouri, 544 U.S. 622, 635 (2005) (quoting Chapman, 386 U.S. at
24); Wall v. State, 184 S.W.3d 730, 746 n.53 (Tex. Crim. App. 2006). In the context of
the denial of physical confrontation, the harm analysis “cannot include consideration of
whether the witness’ testimony would have been unchanged, or the jury’s assessment
unaltered, had there been confrontation” because “such an inquiry would obviously
involve pure speculation.” Coy, 487 U.S. at 1021–22. Instead, harm must be determined
based on “the remaining evidence.”18
b. Analysis
According to the court of appeals, the Van Arsdall harmless-error test applies.
Haggard, 2019 WL 2273869, at *7 (citing Shelby v. State, 819 S.W.2d 544, 546 (Tex.
Crim. App. 1991) (relying on Delaware v. Van Arsdall, 475 U.S. 673, 684 (1986))). In
Van Arsdall, the Supreme Court held that improper limitations on cross-examination that
violate the Confrontation Clause are reviewed for constitutional harm and that the
18
While a reviewing court cannot speculate as to what would have happened had there
been a face-to-face confrontation, it should examine the testimony that the witness actually gave
when determining whether there is a reasonable likelihood that the error affected the judgment of
the jury.
Haggard–18
analysis begins with the presumption that the damaging potential of the cross-examination
was fully realized. The Court then listed several factors to consider,
[(1)] importance of the witness’ testimony in the prosecution’s case, [(2)]
whether the testimony was cumulative, [(3)] the presence or absence of
evidence corroborating or contradicting the testimony of the witness on
material points, [(4)] the extent of cross-examination otherwise permitted,
and, of course, [(5)] the overall strength of the prosecution’s case.
Van Arsdall, 475 U.S. at 684.
The Van Arsdall presumption is inappropriate here where cross-examination is not
at issue.19 But we think the factors set out by the Supreme Court are helpful in
determining harm (except for the fourth one asking to what extent cross-examination was
otherwise permitted). Ultimately, however, any circumstance apparent in the record that
logically informs the harm issue should be considered. Snowden v. State, 353 S.W.3d 815,
822 (Tex. Crim. App. 2011).20
19
Some courts seem to have taken the Court’s citation to Van Arsdall in Coy to mean that
Van Arsdall controls the harm analysis for the denial of face-to-face confrontation. That is not
how we read Coy. We understand the Court to have cited Van Arsdall to show that the denial of
face-to-face confrontation, like impermissible limits on cross-examination, is susceptible to a
harmless-error analysis. In other words, neither the denial of cross-examination nor the denial of
face-to-face confrontation under the Confrontation Clause are structural errors.
20
Haggard argues that the court of appeals also erred because it undertook a legal-
sufficiency analysis, not a denial-of-face-to-face-confrontation harm analysis. Haggard, 2019
WL 2273869, at *7. In particular, he points to the court of appeals’s comment that the
inculpatory evidence minus DeVore’s testimony was “overwhelming.”
It is true that we have used the word “overwhelming” when referring to the sufficiency of
the evidence, but also when determining whether a defendant was harmed by a constitutional
violation. When we use “overwhelming” in a legal-sufficiency analysis, we mean that the issue is
not a close call, i.e., the record evidence is more than sufficient for a rational jury to reasonably
find each essential element of the offense beyond a reasonable doubt. But when we use the word
“overwhelming” in a constitutional harmless-error analysis, we are referring to the diminished
Haggard–19
There is another issue that the court of appeals should have addressed in its
analysis but did not: whether evidence admitted through the complained-of witness must
also be excluded from the analysis?21 The issue is important here because, without
DeVore to prove-up the chain of custody of the SANE kit and its contents, the highly
incriminating DNA evidence would have been inadmissible. We think it must be
excluded under Coy. If it is improper to speculate about how a witness might have
testified had there been physical, face-to-face confrontation, it is also improper to
speculate about whether evidence admitted through the witness would have been admitted
at a hypothetical trial at which the witness did not testify. Engaging in such speculation
could lead to the exact problems that the Supreme Court sought to avoid in Coy.
For instance, even if we speculate that the State might have known another
possible way to get the SANE kit, its contents, and the DNA evidence admitted, that
changes the analysis from analyzing harm in the trial that actually happened to assessing
harm at a theoretical trial.22 We would also have to speculate about whether the witness
would have given the necessary testimony to admit the evidence—often an attorney
probability that a constitutional error materially affected the deliberations of the jury.
“Overwhelming” evidence of guilt will always be legally sufficient to support a conviction, but
constitutional error could be harmful even though the evidence of guilt is “overwhelming.” The
distinction is important.
21
Haggard made this argument to the court of appeals, but it appears to have overlooked
it.
22
We note that the prosecutor told the judge that the only way he knew to get the SANE
kit and its contents admitted was through DeVore’s testimony.
Haggard–20
believes that a witness will say one thing, but he says another—and whether the judge
would have made the proper ruling at the hypothetical trial.
CONCLUSION
Having clarified the harm analysis for the denial of face-to-face confrontation, we
reverse the judgment of the court of appeals and remand the cause to that court to reassess
whether Haggard was harmed.
Delivered: December 9, 2020
Publish