Shapiro, P.J.

Bill Schuette, Attorney General, Aaron D. Lindstrom, Solicitor General, Carol A. Siemon, Prosecuting Attorney, Joseph B. Finnerty, Appellate Division Chief, and Elizabeth L. Allen, Assistant Prosecuting Attorney, for the people.

Law Offices of Casey D. Conklin, PLC (by Donald J. Baranski) for defendant.

Before: Shapiro, P.J., and M. J. Kelly and O'Brien, JJ.

Shapiro, P.J.Defendant appeals his jury conviction of third-degree criminal sexual conduct (CSC-III) (incapacitated complainant), MCL 750.520d(1)(c), and fourth-degree criminal sexual conduct (CSC-IV) (incapacitated complainant), MCL 750.520e(1)(c). The trial court sentenced defendant to 30 to 180 months' imprisonment for the CSC-III conviction and 12 to 24 months' imprisonment for the CSC-IV conviction. On appeal, defendant argues that the trial court erred by granting the prosecution's motion to allow the complaining witness to testify while accompanied by a support dog and its handler. We agree, and so reverse and remand for a new trial.

I. FACTS AND PROCEDURAL BACKGROUND

The incident giving rise to the charges occurred early in the morning of May 1, 2016, in complainant's home in a mobile home community. Complainant and defendant both testified at trial, and their statements about the events of the day were generally consistent until the point in time when they prepared to go to sleep. Complainant and defendant, along with two other friends, traveled to a rodeo together in complainant's car. The group had planned to camp nearby that night after the rodeo, but it rained and they decided to drive home instead. Defendant drove the group home in complainant's car and dropped off the other two friends first. When they arrived at complainant's home, it was about 4:30 a.m., so she told defendant that he could sleep on an air mattress at her home that night and that she would drive him to his house when they woke up. Defendant and complainant had known each other for some time as friends. They had never spent the night together or had any sexual contact with each other, although they both agreed that they had been flirting during the day. After entering the complainant's home, they discovered that the air mattress had been left in the car and because it was raining, neither wanted to go back outside to get it. Complainant answered a phone call and sat on her bed when conversing. During that time, defendant lay down with his head on her upper leg, and complainant did not object. From this point in the chronology, the complainant's testimony and defendant's testimony differ significantly.

Complainant testified that she told defendant that he could sleep with her in her queen-sized bed but that she would not engage in any sexual contact. She testified that "[w]e agreed that we could—as both adults we could sleep in my bed without being intimate, and I made that very clear to him." She explained that they dressed in separate rooms, and that she was clothed in leggings and a t-shirt with a bra underneath. According to complainant, at about 6:30 a.m., she woke to the sound of her roommate leaving for work, and discovered that defendant was sexually assaulting her. She testified that "[w]hen I woke up to my roommate leaving for work. I was on my back and [defendant] was to the left of me and he had unfastened my bra and his left hand was in my pants." She explained that defendant had pulled down the front of her leggings, that he was penetrating her vagina with his fingers, and that his other hand was underneath her shirt and bra on her breast. Complainant stated that she was "terrified" and that she immediately jumped up, told defendant she needed to go to the bathroom, and went to the end of the hallway to her roommate's room to seek help from Tyler Johnson, her roommate's boyfriend.

Defendant's testimony as to what occurred before they fell asleep differed from complainant's only in that he testified that she did not make any statements about not wanting to have any sexual contact. Defendant testified, however, that he did not expect that there would be any sexual activity and that they changed clothes in separate rooms. Defendant testified that he was awakened by the sounds of the roommate leaving, and that when he awoke; he found that complainant had wrapped herself around him, i.e. she had "cuddled up" to him. According to defendant, complainant's leg was across his waist and her head was on his chest. He stated that he attempted to roll her off him by pushing her hip and thigh, but that she "snuggled" back. He admitted that he kissed her on the forehead. Defendant testified that he then fell back asleep. He testified that there had been no sexual contact, and he stated that the complainant's behavior after she awoke was a complete surprise and that he did not understand it. On cross-examination, he agreed that he did not have the complainant's consent to do anything sexual, and he repeated that he had not done so. He was also asked about statements he had made to the officer that the prosecution argued were inconsistent with his testimony. Defendant agreed that he told the officer that he touched complainant's butt and thigh and explained that this occurred while he was attempting to move her off him. When asked whether he had admitted to the officer that he had touched complainant's belly and back under her shirt, defendant denied making such a statement. Defendant also testified that he could not offer a reason why the complainant would make up the charges against him.

Johnson, complainant's housemate's boyfriend, testified that when complainant entered his room that morning, she was so upset that he could barely understand her. Because Johnson's phone was dead, he and complainant walked to the park manager's home to call the police. The manager described complainant's emotional state as "absolutely hysterical to the point I couldn't understand her." He concluded that complainant was saying that there was a dead body in her bed and he informed the police of this in his 911 call. The manager testified that he then walked back to complainant's house just as defendant was coming out the door and he told defendant that he should come with him because he had been reported as dead. Finally, he testified that when he and defendant reached his house, he asked defendant if he had done something wrong and defendant responded, " ‘Yea, I did.’ " The police officers, who testified, agreed that complainant was very upset when they arrived. One officer testified that "[s]he was shaking and her breathing was to the point where [she was] almost hyperventilating."After giving her statement to the police, the complainant went to Sparrow Hospital for a sexual assault examination. The sexual assault nurse examiner found no injury to the complainant's vaginal area. A DNA test of the surface of her breast revealed male DNA, but upon testing, it was determined that the DNA did not belong to defendant. Before the start of trial, the prosecutor and defense counsel informed the court of these results, and the parties agreed that it was not necessary to present a full chain of custody for the DNA.

It appears uncontested that the defense counsel was not informed before trial that the prosecution intended to have complainant testify while accompanied by a support dog and the animal's handler. It was first raised during voir dire when the prosecutor asked potential jurors about the possibility of a support animal in the courtroom while the complainant testified. On the following trial day, before opening statements, the prosecution raised the issue again. When the judge learned that in addition to the dog, a handler would have to be present next to the witness stand to hold the dog's leash, she expressed concern about whether there was evidence of a "necessity for that support animal." Defense counsel objected to the use of a support dog, noting that there had to be "some basis for the need for that animal to be here. You know, I usually experience this in kid's cases or something of that nature, and we're dealing with an adult woman here. I'm not sure if there's some special exception that gives rise to the need for this, but at this point I would ask that she testify solely on her own." The trial court, citing People v. Johnson , 315 Mich. App. 163, 889 N.W.2d 513 (2016), stated that "there should be some kind of a showing of a need that it would promote the expeditious and effective ascertainment of the truth." The prosecution responded:

Judge, I think my response to that is it will limit her emotional outbursts when she's testifying. This is a victim who has been teary eyed multiple times when I've spoken to her, without even getting into, you know, the facts of what occurred to her and the actual testimony that she will be giving. We did a trial preparation meeting at my office last week where Preston, the dog, was present. She was less emotional with him in the room. She indicated she felt more comfortable and that this is something she wants. I think that it—it would be a benefit to both sides to have her control her emotions through the use of the support dog.

The trial court then granted the prosecution's request, concluding: "That's sufficient for this Court in that it will limit her emotional display on the stand. I agree that that could even be beneficial to the Defendant. And that there's already been, sort of, a trial run with the dog and it's been a successful one at that. So, I think that's a sufficient basis for her to use the support animal while testifying."

When instructing the jury at the outset of the complainant's testimony and, again, at the end of the trial, the court instructed the jury that it must "not allow the use of a support animal to influence your decision in any way" and that it "should not consider the witness's testimony to be any more or less credible because of the animal's presence." Closing arguments focused on the credibility of the complainant and defendant, and whom the jury should believe. The jury returned, after several hours, and advised the court that they had reached a verdict as to the CSC-IV charge, but not as to the CSC-III charge. The court gave them the deadlocked jury instruction; they resumed deliberations and, sometime later, returned a verdict of guilty on both counts.

II. ANALYSIS

Defendant contends that the trial court erred by allowing the complainant to testify while accompanied by a support animal and its handler. In making its ruling, the trial court relied on Johnson , 315 Mich. App. 163, 889 N.W.2d 513. In that case, the defendant was accused of sexually assaulting his six-year-old niece, and the trial court permitted a support dog to be with the six-year-old and her ten-year-old brother during their testimony. Id . at 171-172, 889 N.W.2d 513. The defense appealed on the ground that allowing the support dog was improper. Id . at 173, 889 N.W.2d 513. The Johnson Court affirmed the conviction, finding no error and concluding, as a matter of first impression, that allowing the use of a support dog is within the authority of a trial court and is not "inherently prejudicial." Id . at 181, 889 N.W.2d 513. Johnson also held that the statute governing the use of a support person, MCL 600.2163a(4), did not apply to support animals. Id . at 175-176, 889 N.W.2d 513.

The first question before us, therefore, is whether Johnson controls our decision. We conclude that it does not. In reaching this conclusion, we are cognizant of the fact that the language used in Johnson sweeps with a broad brush and that the opinion contains language which, if one does not consider the facts of the case, could seem to be dispositive here. However, we conclude that there is a fundamental difference between allowing a support animal to accompany a child witness, as in Johnson , and allowing the animal to accompany a fully abled adult witness, as in the instant case. Accordingly, we conclude that Johnson's holding was tied to the facts presented, i.e., the use of a support animal during a child's testimony. We do so for several reasons.

First, allowing support animals for fully abled adults would be unprecedented, not only in Michigan, but apparently nationwide. Johnson cites many out-of-state cases in support of its conclusion, but each of the cited cases involved a witness who was either a child or an adult with a developmental disability. Indeed, despite our efforts, we have been unable to find a case in any jurisdiction allowing the use of a support animal or a support person when the witness is a nondisabled adult. Therefore, Johnson's observation that "other jurisdictions have upheld this procedure as part of a trial court's inherent authority to control the courtroom," id . at 171, 889 N.W.2d 513, while broadly stated, rests solely on cases involving children or developmentally disabled adults.

Second, any reliance placed on MCL 600.2163a in Johnson was inapposite because that statute, by its terms, applies only if the witness is a child or a developmentally disabled adult. Johnson noted that this statute grants a trial court the authority to allow a support person or take other action to protect a witness, such as using a screen. Johnson , 315 Mich. App. at 176-177, 889 N.W.2d 513. However, by its terms, MCL 600.2163a limits this authority to cases in which the witness is either "a person under 16 years of age," "a person 16 years of age or older with a developmental disability," or a "vulnerable adult." MCL 600.2163(1)(f). The statute defines "developmental disability" as "a condition that is attributable to a mental impairment or to a combination of mental and physical impairments...." MCL 600.6123a(1)(c). The term "vulnerable adult" is defined in other statutes as: "an individual age 18 or over who, because of age, developmental disability, mental illness, or physical disability requires supervision or personal care or lacks the personal and social skills required to live independently" or a person who is "unable to protect himself or herself from abuse, neglect, or exploitation because of a mental or physical impairment or because of advanced age." There is nothing in the record to suggest that the complainant was developmentally disabled, mentally ill or physically disabled so as to require supervision, or that she was unable to live independently. MCL 600.2163a; MCL 750.145m(u)(i). Accordingly, MCL 600.2163a provides no authority for the situation presented in this case.

Third, Johnson refers to the trial court's inherent authority over the courtroom. We agree that this authority is broad and that the practice of allowing child witnesses to have a support animal or person is common enough that it is comparable to the examples offered in Johnson of that authority. However, allowing a nondisabled adult witness to have a support animal or person with them during testimony is a different circumstance. Given that we cannot find a single prior example of any court allowing such a procedure, we hesitate to conclude that it is simply a matter of a trial judge's inherent authority. If that were so, one would presumably be able to find prior examples of that authority being exercised somewhere.Fourth, we note that the Johnson decision repeatedly and properly notes the tender ages of the children testifying. Moreover, we find nothing in the opinion, despite its broad language, that indicates that the Court even considered whether its holding would apply in the setting of a fully abled adult. As already discussed, none of the cases relied on in Johnson considered the issue in that context.

Lastly, we note that Johnson made no mention of allowing an animal handler to also accompany the witness during testimony. In the instant case, the complainant was accompanied not only by the dog, but also by the handler, who is a human being, which, unlike the situation in Johnson , triggers the requirements of MCL 600.2163(4) even if the witness is a child.

Even assuming a trial court had the inherent authority to allow such a procedure, we would not approve its use if the basis for it was simply that doing so would allow the witness to be "more comfortable" or because "this is something she wants." Nor are we convinced that allowing a support animal or person so that the witness would be better able to "control her emotions" necessarily aids the truth-finding process. This is particularly so in this case, where the prosecution presented evidence from four different witnesses that the complainant was hysterical, shaking, and barely able to speak following the alleged assault and argued to the jury that her emotional reactions were evidence of defendant's guilt. In that context, it was particularly improper to allow a comfort dog to help the complainant "control her emotions" while testifying. If the adult complainant's emotional state constitutes evidence of guilt, the jury is entitled to evaluate her emotional state uninfluenced by outside support, not only as it pertains to her own credibility, but also to determine the weight to be given to testimony by others who described her emotional state.

For all these reasons, we conclude that allowing the complainant to be accompanied by a support dog during her testimony was error.

Although the prosecution does not ask that we consider whether any such error was harmless, we have done so, and we conclude that the error was not harmless because it undermined the reliability of the verdict. People v. Feezel , 486 Mich. 184, 192, 783 N.W.2d 67 (2010). There were no witnesses to the actual events other than defendant and complainant, and the only forensic evidence consisted of a different male's DNA. Thus, the case turned almost exclusively on the jury's evaluation of their respective credibility. Secondarily, it turned, in part, on the weight given by the jury to the testimony concerning complainant's emotional state after leaving her bedroom that morning. These determinations were likely affected by the use of the support dog and handler. A juror can readily accept that a child might need support simply to be in a courtroom to be able to answer questions. With a fully abled adult, a juror is far more likely to conclude that the reason for the support animal or support person is because the complainant was traumatized by the actions for which the defendant is charged.

In sum, we conclude that a fully abled adult witness may not be accompanied by a support animal or support person while testifying. Indeed, were we to rule that a fully abled adult may be accompanied by a support dog or person simply because they will be "more comfortable," we would unlock a door that we have great hesitation about opening. At a minimum, this unprecedented change, if adopted, should be made by legislation, court rule, or a decision of our Supreme Court.

Reversed and remand for a new trial. We do not retain jurisdiction.

M. J. Kelly, J., concurred with Shapiro, P.J.

O'Brien, J. (concurring in part and dissenting in part ).

I agree with the majority that the trial court erred in this case by allowing the victim to use a support dog while testifying. However, I believe that the error was harmless. Therefore, I respectfully dissent in part.

For preserved, nonconstitutional errors, reversal is not required unless the defendant establishes that it is more probable than not that the error was outcome determinative. People v. Lukity , 460 Mich. 484, 495, 596 N.W.2d 607 (1999). This standard does not "require actual innocence, but, rather, it should be viewed as a legislative directive to presume the validity of verdicts and to reverse only with respect to those errors that affirmatively appear to undermine the reliability of the verdict." People v. Mateo , 453 Mich. 203, 211, 551 N.W.2d 891 (1996) ; see also MCL 769.26. Thus, "[a]n error is deemed to have been ‘outcome determinative’ if it undermined the reliability of the verdict." People v. Elston , 462 Mich. 751, 766, 614 N.W.2d 595 (2000). When determining whether an error was harmless, the focus is on the nature of the error "in light of the weight and strength of the untainted evidence." Mateo , 453 Mich. at 215, 551 N.W.2d 891.

I agree with the majority that this case—like the vast majority of criminal sexual assault cases—had "no witnesses to the actual events other than’’ defendant and the victim and, therefore, largely came down to "their respective credibility." However, unlike the majority, I believe that the effect—if any—of the support dog on the victim's and defendant's credibility was harmless in light of the untainted evidence that bolstered the victim's credibility and damaged defendant's.

The manager of the trailer park where the victim lived, Joshua Kaimon, testified that, shortly after the assault occurred, defendant admitted to him that he had done something wrong. Specifically, Kaimon testified that, on the morning of the incident, he was outside his trailer with defendant waiting for the police to arrive and asked defendant if he did "something [he] shouldn't have," to which defendant replied, " ‘Yea, I did.’ " Officer Aaron Reynolds, who interviewed defendant on the morning of the assault, testified that defendant initially denied that anything occurred while he was sleeping with the victim in her bed. However, according to Officer Reynolds, defendant later changed his story and admitted that, while the victim was sleeping, he attempted to kiss her, touched her butt, placed his hand on her upper thigh, and placed his hand underneath the victim's shirt and touched her belly and back area. Officer Reynolds also testified that defendant told him that he had an erection "while they were cuddling." The testimony of Officer Reynolds, besides damaging defendant's credibility, provided substantive evidence to convict defendant of fourth-degree criminal sexual assault (CSC-IV), MCL 750.520e(1)(c), if the jury found that defendant's touching of the victim's buttock while she was asleep was intentional and for a sexual purpose, see MCL 750.520a(f) and (q). The majority does not contend that Officer Reynolds's testimony was tainted by the victim's use of a support dog. Therefore, even under the majority's analysis, defendant's CSC-IV conviction should be upheld. See Mateo , 453 Mich. at 215, 551 N.W.2d 891.

In contrast to defendant's changing story, the victim's story remained unchanged: she testified at trial that she woke up to defendant's fingers inside her vagina and another hand on her bare breast. The officer who interviewed the victim following the incident, as well as the sexual-assault nurse examiner who examined the victim, testified that, when they spoke with the victim on the morning of the incident, she told them that she woke up to defendant's fingers inside her vagina and his other hand on her breast. Evidence of the victim's unchanging story bolstered her credibility, while evidence of defendant's incriminating admissions and changing story damaged his credibility. All this evidence was untainted by the trial court's error of allowing a support dog to accompany the victim.

The majority's harmless-error analysis does not appear to assess the effect of the trial court's error "in light of the weight and strength of the untainted evidence." Mateo , 453 Mich. at 215, 551 N.W.2d 891. Instead, it rests on the following analysis:

A juror can readily accept that a child might need support simply to be in a courtroom to be able to answer questions. With a fully abled adult, a juror is far more likely to conclude that the reason for the support animal or support

person is because the complainant was traumatized by the actions for which the defendant is charged.^{[ ]}

When reviewing the information—more accurately, the lack of information—placed before the jurors regarding the support dog, the majority's holding becomes unduly speculative. The support dog was barely mentioned during the course of the three-day trial. The first time it was mentioned was during voir dire when the prosecutor told jurors that the victim may be accompanied by a support dog. The prosecutor asked jurors not to draw any conclusions from the dog's presence, and all jurors confirmed that they could ignore the dog. The prosecutor did not discuss the dog any further.

The next time that the dog was mentioned to jurors was before the victim testified. At that time, the trial court told jurors that the victim would be accompanied by a support animal and gave the following instruction:

You should disregard the [support] animal's presence and decide the case based solely on the evidence presented. You should not consider the witness's testimony to be any more or less credible

because of the animal's presence. You must not allow the use of a support animal to influence your decision in any way.

The support dog was not mentioned at any point during the victim's testimony, no other witness was asked about or mentioned the support dog, and neither party referred to the support animal during closing arguments. The dog was only mentioned one other time: during final jury instructions, the trial court repeated its instruction that the jurors were to ignore the support animal and not draw any conclusions from its presence.

On this record, I see no reason to make the leaps necessary to find that a juror was "likely to conclude that the reason for the support animal ... [was] because the complainant was traumatized by the actions for which the defendant [was] charged." No one gave any reason to jurors for why the support dog was accompanying the victim. In fact, no one even explained to jurors what a support animal was. Thus, the majority's conclusion assumes (1) that a juror had general knowledge about why someone may use a support animal and (2) what the juror was "likely to conclude" from the animal's presence. With respect to this second point, this Court has recognized that the use of a support dog "does not give rise to primarily prejudicial inferences, as it is possible for the jury to make a wide range of inferences from the use of this procedure that are unrelated to defendant." People v. Johnson , 315 Mich. App. 163, 180, 889 N.W.2d 513 (2016). Yet, without explanation, the majority concludes that "a juror is far more likely to conclude that the reason for the support animal" is related to defendant. The majority's conclusion also assumes that (1) the juror disregarded the trial court's instructions, despite the general rule that jurors are presumed to follow those instructions, People v. Graves , 458 Mich. 476, 486, 581 N.W.2d 229 (1998), and (2) the juror lied when he or she confirmed during voir dire that they could ignore the dog's presence. In my opinion, the majority's conclusion is "unduly speculative" and does not warrant finding that the error in this case was not harmless. Id . at 485-486, 581 N.W.2d 229 (recognizing that an error is harmless if a conclusion to the contrary is "unduly speculative").

In sum, the only error was the trial court’s allowing a support dog to accompany the victim. This dog—which was barely mentioned throughout trial—did not, by its mere presence, add credibility to the victim's claim of a traumatic experience because none of the jurors knew why the dog accompanied the victim. More importantly, any credibility that the support dog's presence may have added to the victim—or taken away from defendant—was minimal in light of the significant untainted evidence that damaged defendant's credibility and supported the victim's. Considering the untainted evidence, it is not more probable that, had the victim testified without the aid of a support dog, a different outcome would have resulted. See Lukity , 460 Mich. at 495, 596 N.W.2d 607.

Because I do not believe that the error affected the reliability of the jury verdict, I would leave the jury's verdict untouched.