SCOW Docket: Expert witnesses permitted to weigh in on witness credibility ... sometimes

Note: We are crunching Supreme Court of Wisconsin decisions down to size. The rule for this is that no justice gets more than 10 paragraphs as written in the actual decision. The “upshot” and “background” sections do not count as part of the 10 paragraphs because of their summary and necessary nature. We’ve also removed citations from the opinion for ease of reading but have linked to important cases cited or information about them. Italics indicate WJI insertions except for case names, which are also italicized. 
​

​The case: State of Wisconsin v. Jobert L. Molde

Majority: Justice Brian J. Hagedorn delivered the unanimous opinion of the court (10 pages)

​Concurrence: Justice Jill J. Karofsky (5 pages)

The Upshot
Under the Haseltine rule, witnesses may not testify that they think another witness is telling the truth. Vouching for the credibility of another witness is impermissible under the rules of evidence because it invades the province of the trier of fact—here, the jury. The question in this case is whether an expert witness violated the Haseltine rule when she testified that only around one percent of child sexual assault disclosures are false, but did not offer an opinion on whether the victim in this case was telling the truth. We conclude she did not. We hold that statistical evidence alone on the likelihood of false reports does not violate the Haseltine rule. The defendant here alleges his counsel was constitutionally deficient for not raising a Haseltine objection to this testimony. Because such an objection would have failed, the defendant’s claim for ineffective assistance of counsel fails as well.

Background
The issues in this case arose following allegations that, sometime between January 2011 and January 2012, Lauren was sexually assaulted by her father, Jobert Molde. This came to light in 2017 when Lauren—now age thirteen—attempted suicide. After Lauren’s claims were investigated, Molde was charged with one count of first-degree sexual assault of a child who had not attained the age of twelve and one count of incest with a child. The crucial evidence against Molde was Lauren’s in-court testimony recounting the assault and a recording of her forensic interview.

The circuit court granted the State’s motion to have the nurse practitioner who conducted the forensic interview testify as an expert at trial. However, she was unavailable, and the circuit court permitted Dr. Alice Swenson—a licensed child abuse pediatrician who supervised Lauren’s forensic interview and examination—to testify instead. The record is unclear about what her supervision entailed other than it was in real time; but Dr. Swenson did not personally conduct an evaluation of Lauren. Dr. Swenson testified about her background and work as a licensed child abuse pediatrician, how child forensic interviews tend to proceed, what sort of evidence they look for in a physical forensic exam, background about how children’s memories work, Lauren’s admission to the hospital, and the possibility of intercourse between an adult and child. She did not testify about Lauren’s truthfulness or how likely it is that Lauren was telling the truth during the interview.

After Dr. Swenson’s testimony, one juror submitted two questions, which the circuit court previously instructed was permissible. Following a sidebar, Molde’s counsel did not object to the questions being read to the witness:

THE COURT: Doctor, it says how frequent is it for children to make up a story of sexual abuse?

THE WITNESS: False disclosures are extraordinarily rare, like in the one percent of all disclosures are false disclosures.

THE COURT: Second part of that is why would they do that?

THE WITNESS: I don’t think I really have an answer to that.

Molde’s attorney did not object or otherwise challenge Dr. Swenson’s answers. The court then permitted Molde’s counsel to ask a follow-up question:

[MOLDE’S COUNSEL]: Are there particular studies that have been conducted regarding the reporting of false accusations?

THE WITNESS: There are that I’ve read, yes. I don’t know the names of them off the top of my head.

The trial proceeded and the jury found Molde guilty on both counts. Following his conviction, Molde moved for postconviction relief. He contended that his trial counsel should have objected to Dr. Swenson’s testimony as impermissible vouching, and this failure constituted ineffective assistance of counsel. The circuit court denied the motion, in part because Dr. Swenson did not comment “on the credibility of the victim in this case as to whether she was telling the truth or not.”

Relying on its prior published decisions, the court of appeals held that Dr. Swenson’s testimony constituted impermissible vouching, and that Molde’s attorney was constitutionally ineffective for failing to object. The state petitioned for review, which we granted.

The Guts
In Wisconsin, the trier of fact—often a jury—is entrusted with the duty to make factual determinations at trial. As part of that role, the jury must decide for itself whether to believe a witness’s testimony in whole, in part, or not at all.

In State v. Haseltine, the court of appeals considered the testimony of a psychiatrist in a case likewise involving a father’s sexual assault of his daughter. The psychiatrist, who was qualified as an expert, testified regarding typical patterns of behavior for victims of incest. But the court also permitted the psychiatrist to offer his professional opinion that the victim fit the typical case, and that he had “no doubt whatsoever” that the father sexually assaulted his daughter. The court of appeals held that this was error, and laid out some basic principles that continue through our cases today.

First, the expert’s opinion that the victim was telling the truth went too far. The foundation for this is WIS. STAT. § 907.02 which then, as now, states in relevant part that an expert witness’s testimony must “assist the trier of fact.” Because the jury is the ultimate arbiter of credibility, “[n]o witness, expert or otherwise, should be permitted to give an opinion that another mentally and physically competent witness is telling the truth.” This principle—that vouching for the credibility of another witness is impermissible—has since become known as the Haseltine rule.

Even so, expert testimony that helps the jury assess credibility or understand the victim’s testimony is permitted. The court explained that expert evidence regarding why incest victims might delay in reporting or recant accusations could aid the jury, which might otherwise “regard such behavior as an indication that the victim was not telling the truth.”

Thus, the key principle is this: expert testimony may assist the jury in determining the credibility of another witness, but it may not supplant the jury’s role by opining on the witness’s credibility. To be sure, any expert testimony should be properly admitted under the appropriate rules. And of course, it is subject to thorough vetting and cross-examination by opposing counsel. The jury can then weigh the credibility of the expert testimony and consider that along with other evidence to determine the credibility of the victim and other testifying witnesses. This rule preserves the jury’s exclusive role as the trier of fact, while permitting the jury to consider all relevant information in making its ultimate determinations.

…

(T)he main question is whether the testimony assists the jury in assessing the credibility of witnesses, or whether it functionally usurps the jury’s fact-finding role. Generally, expert testimony describing typical behavior by those in similar circumstances can serve to assist the jury. Provided it meets the other evidentiary requirements for admissibility, generalized evidence and personal observations do not, by themselves, run afoul of the Haseltine rule. Witnesses cross the line, however, when they take the extra step of implicitly or explicitly opining on whether the complainant is telling the truth. This usurps rather than assists the jury. Whether this has occurred will necessarily be dependent on the facts of a given case.

The main question presented here is whether expert testimony regarding the likelihood of false reporting, particularly when the evidence demonstrates that false reports are extremely rare, constitutes impermissible vouching. Molde says it does. The court of appeals agreed in a recent case where it held that expert testimony impermissibly vouched for the victim’s credibility when two experts testified that 99.2% and 99.33% of victims are truthful. It reasoned in part that those percentages amounted to near-mathematical certainty “that false reporting simply does not occur.” This violated Haseltine, the court of appeals ruled, because the jury would inevitably understand this testimony as a statement that the victim was telling the truth.

Applying this holding (as it was bound to do), the court of appeals in this case held that Dr. Swenson’s testimony that false reports in these circumstances happen around 1% of the time—meaning 99% of child sexual assault reports are true—violated Haseltine. We see it differently.

Statistical evidence alone is precisely the kind of generalized evidence that might assist the jury, not usurp its role. It does not matter that typical behavior helps one side or another. Statistical evidence by itself does not tell the jury which category—truthful or untruthful—a particular witness belongs to. The jury still must assess the credibility of the statistical evidence and that of the expert, and then weigh that along with the other evidence in the case. The Haseltine rule is not violated simply because generalized or typical evidence strongly suggests a complainant is telling the truth. A Haseltine violation requires the “extra step” of the expert actually opining on the truthfulness of the complainant.

We stress that our conclusion does not mean all such evidence should be admitted or is impervious to attack. The circuit court still must determine such evidence is admissible—including that the expert is qualified and that her conclusions are “the product of reliable principles and methods” that have been applied “reliably to the facts of the case.” Expert testimony is also subject to other normal evidentiary rules such as the exclusion of evidence that, although otherwise admissible, raises too great a danger of unfair prejudice. Molde did not challenge the admissibility of Dr. Swenson’s testimony on these grounds, nor does he raise any similar challenge here. In addition, defendants can always challenge the accuracy of statistical evidence and otherwise attack the credibility of expert witnesses through cross-examination, competing experts, and other means. The adversarial process is designed to assist the trier of fact. We are not persuaded that statistical evidence which strongly supports a complainant’s story necessarily interferes with the fact-finder’s prerogative to determine that witness’s credibility.

​Concurrence
I write separately to elevate the voice of Lauren, the victim in this case whose courage and perseverance allowed her to overcome the “herculean task of reporting sexual abuse.” Lauren’s father, Jobert Molde, sexually assaulted her when she was eight or nine years old. Lauren found her voice five years later, after a suicide attempt, when she reported the crime. Lauren found her voice again during her forensic interview when she relayed details of the assault. And at trial, in front of a jury of strangers, she found her voice yet again to testify about the abuse. Lauren managed to state, in open court, that after her father told her “to be his big girl for daddy,” he “had sex” with her by putting his penis in her vagina, which “hurt.” The jury believed her.

The court of appeals did not. The court of appeals overturned Molde’s conviction of first-degree sexual assault of a child, relying in large part on its determination that Lauren lacked credibility. The court reached this conclusion because “[t]he sexual assault allegation was not independently corroborated by other evidence; there was no physical evidence; there was only one sexual assault that occurred during a one-year period roughly four to five years prior to Lauren’s accusation; and some aspects of Lauren’s story changed over time.”

These assertions are as disconcerting as they are misguided. They are predicated upon damaging and victim-blaming misperceptions. Disregarding a child victim’s testimony because of delayed reporting, small variations in her narrative, and most alarmingly, the total number of assaults she reported, defies what we know about how child sexual assault victims behave and report.

…

First, delayed disclosure and a lack of corroborating evidence are the norms in child sexual assault cases.  (Emphasis in original.) The court of appeals deemed Lauren less credible because she reported the incident four or five years after Molde assaulted her, and there was no corroborating evidence. Delayed disclosures are not uncommon. In fact, “70–75% of child sexual abuse victims wait five years or more before disclosing the abuse.” Studies show that only about 30% of those who have experienced child sexual abuse disclosed it during childhood. There are many reasons for these delays, including “an inability to recognize or articulate sexual abuse, an uncertainty about which adults are safe, a lack of opportunities to disclose, fear of not being believed, trauma … and institutional power dynamics.”

…

Second, victims often cannot perfectly recall and recite their trauma. (Emphasis in original.) The court of appeals deemed Lauren less credible for omitting a detail from her forensic interview, changing the timing of a conversation with her younger sister, and being unsure whether Molde was clothed when she first entered his bedroom, even though the rest of her account remained notably consistent. These minor inconsistencies do not necessarily suggest a credibility problem. Rather, they illustrate how most victims report after enduring a traumatic event.

Finally, the number of assaults reported has absolutely nothing to do with credibility. (Emphasis in original.) The court declared that Lauren lacked credibility because she suffered only one assault.

It is difficult to even begin to respond to such a baseless statement, especially because, as noted above, most sexual assault victims underreport, or never report, their abuse. It would be completely unreasonable to require robbery victims to suffer two or more robberies before believing them. Why place such a burden on child sexual assault victims? And even if such a credibility threshold exists—to be clear it should not—is there a magic number of sexual assaults that a child must endure in order to be credible? Is it two or three or ten? One can only imagine the compounded chilling effect this elevated credibility standard would have on the already delayed and limited reporting of child sexual assault.
​
Lauren defied the odds in reporting her abuse. The court of appeals discredited her for reasons that either fail to account for commonplace behaviors of child sexual assault victims or ignore logic and common sense. Child sexual assault victims must overcome near-insurmountable barriers to reporting abuse and achieving justice. When these brave children speak, courts must ensure they are heard.

---