Haggard, James Ray

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, headhair 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