2. Background and Procedural History
The investigation and DNA evidence
M.B. began living with Welch and her mother in 2017. By 2021, the children lived with them full-time, visiting their father on alternating weekends. Because their mother frequently worked overnight, Welch often cared for the children in her absence.
After police learned of the pregnancy, officers collected DNA samples from M.B., H.B., M.B.’s father, and Welch. Additional samples were later collected from M.B. and H.B. The Vermont Forensic Laboratory analyst testified that the genetic results were seventy-two trillion times more likely if Welch was H.B.’s father than if the father was an untested, unrelated, randomly selected man.
Before trial, Welch sought to limit the State to one set of samples and the corresponding reports and custody evidence, arguing that additional samples were needlessly cumulative. The trial court denied that motion, accepting the State’s explanations that the evidence demonstrated a thorough investigation, established unstipulated chains of custody, and supported proof that a sexual act occurred. The appellate analysis, however, focused on specific hearsay objections to documents and allele tables, rather than separately resolving the earlier cumulative-evidence argument.
The defense concerning sleep-related behavior
Welch testified that he sometimes learned he had engaged in sexual activity with an adult partner but had no memory of it. He also described finding M.B. in his bed and maintained that he did not recall intercourse with her. His mother testified about a family history of sleep disturbances.
A physician specializing in sleep medicine explained that “sexsomnia” is a rare parasomnia involving sexual behavior arising from sleep. He identified sleep deprivation and obstructive sleep apnea as potentially relevant factors, but he neither examined Welch nor diagnosed him with sexsomnia. Asked whether it could plausibly explain the events, he described it as “a consideration.” The testimony therefore supplied a possible explanation, not a clinical finding that Welch experienced such an episode.
The record also included earlier allegations by M.B. against Welch, which she later told her aunt she had fabricated. Welch relied on that history when explaining his concerns about what she might say. The appellate court did not independently resolve the truth of those earlier allegations.
The recorded calls
The State introduced portions of calls between Welch and M.B.’s mother following his arrest. Welch discussed the prosecution’s DNA evidence, whether M.B. would testify, keeping her happy, and what would happen if she maintained the same account. He also made statements concerning the pregnancy, the baby, and his continuing relationship with M.B.
Welch offered innocent explanations for those statements. For example, he said his expressions of affection were parental, his concern about keeping M.B. happy related to her mental health and possible false allegations, and his discussion of giving H.B. his surname concerned adoption.
The jury convicted Welch. The trial court denied his post-trial motion for acquittal and a new trial, and he appealed.
4. Analysis
A. Precedents Cited and Their Influence
Review of evidentiary rulings
State v. Bergquist, 2019 VT 17, ¶ 22, 210 Vt. 102 supplied the deferential standard governing appellate review of evidentiary decisions. Its significance is institutional: trial judges ordinarily have latitude to evaluate evidence, and an appeal is not a fresh determination of admissibility.
State v. Lyddy, 2025 VT 1, ¶ 13, 220 Vt. 510 supplied the more specific reversal standard. The defendant must show that discretion was entirely withheld or exercised on clearly untenable or unreasonable grounds, and that the resulting abuse caused prejudice. The court invoked this standard both generally and when rejecting the challenge to the allele tables.
State v. Groce, 2014 VT 122, ¶ 18, 198 Vt. 74 supported harmless-error review because Welch contemporaneously objected to the disputed documents. Preservation allowed appellate examination of the alleged error, but did not eliminate the need to establish that it affected the trial’s outcome.
Harmless error and authentication of physical evidence
State v. Tester, 2009 VT 3, 185 Vt. 241 was central to the document issue. It instructs courts assessing an evidentiary error to imagine a trial without the challenged evidence and consider four factors: the evidence’s importance to the prosecution, its cumulative nature, the extent of cross-examination, and the overall strength of the prosecution’s case.
The same decision also established that chain of custody ordinarily can be demonstrated through evidence that a sample was sealed and labeled when collected and reached the testing technician in that condition. It expressly rejected a requirement that officers provide written documentation to establish adequate custody. These principles defeated Welch’s premise that excluding the documents necessarily would have excluded the DNA results.
Dep't of Soc. Welfare v. Miller, 157 Vt. 92, 96 (1991) supplied the sealed-and-labeled formulation quoted through State v. Tester. Its role was to identify a practical basis for authenticating a tested sample without demanding exhaustive documentary proof of every transfer.
State v. Stevens, 137 Vt. 473, 477 (1979) demonstrated that even a gap in the recitation of custody does not necessarily preclude admission. The controlling concern is reasonable assurance of sample identity, not a perfectly reconstructed custody history.
State v. McAllister, 2008 VT 3, ¶ 10, 183 Vt. 126 reinforced that “reasonable assurance” standard. Together, these authorities directed attention to the reliability established by the witnesses rather than the mere presence or absence of custody paperwork.
State v. Lander, 155 Vt. 645, 645 (1990) (mem.) addressed refreshed recollection. A witness may consult a prior writing and, if memory is successfully refreshed, testify from present recollection. If recollection remains exhausted, admitting the writing itself requires an applicable hearsay exception. This distinction answered Welch’s argument that laboratory witnesses’ references to the disputed documents made those documents indispensable as exhibits.
DNA tables and the source of expert evidence
State v. Walker, 212 A.3d 1244, 1255 (Conn. 2019) supported the treatment of testimony concerning DNA profiles generated by the witness herself as nonhearsay. The Vermont court used it as a comparison because the analyst here personally performed the analyses underlying the tables.
Commonwealth v. McCowen, 939 N.E.2d 735, 756 (Mass. 2010) supplied the contrasting situation. Portions of an allele chart derived from another analyst’s testing were hearsay because they incorporated factual findings of a nontestifying investigator. The distinction made the source of the information—not simply the chart’s scientific appearance—important to the analysis.
Review of jury instructions
State v. Kolibas, 2012 VT 37, ¶ 11, 191 Vt. 474 required consideration of the instructions as a whole: whether they adequately guided the jury without introducing prejudice into deliberations.
State v. Snow, 2013 VT 19, ¶ 8, 193 Vt. 390 required instructions accurately stating the law on every theory fairly supported by the evidence, while recognizing trial-court discretion over wording within legal limits.
State v. Shabazz, 169 Vt. 448, 450 (1999) supplied the ultimate measure of instructional error: whether the instructions undermine confidence in the verdict. The court returned to that measure when concluding that the omitted instruction did not warrant reversal.
Consciousness of guilt and circumstantial proof of intent
State v. Scales, 2017 VT 6, 204 Vt. 137 was Welch’s principal authority. There, a defendant gave police an alternate name when officers sought to execute a warrant. The conduct was ambiguous, the name apparently was used consistently, and little evidence connected the defendant’s knowledge of the warrant to the Vermont crime charged years earlier. The evidence also lacked a connection to a disputed element. The court found its admission erroneous and concluded that refusing a consciousness-of-guilt instruction compounded the prejudice.
The present court rejected the proposition that State v. Scales mandates such an instruction whenever evidence might reveal consciousness of guilt. Its reasoning depended on the particular ambiguity and weak inferential connections in that earlier case.
State v. Perrillo, 162 Vt. 566, 570 (1994) supplied the four-step framework for flight evidence: conduct must indicate flight; flight must indicate consciousness of guilt; that consciousness must concern the charged crime; and it must support an inference of actual guilt. The weaknesses in this inferential chain explain why flight-related conduct may be only marginally probative.
State v. Theriault, 2014 VT 119, ¶ 7, 198 Vt. 625 (mem.) recognized that intent is rarely proved directly and commonly must be inferred from conduct. It supported treating Welch’s case-specific statements as circumstantial evidence of disputed intent rather than automatically assigning them the limited significance associated with flight.
B. Legal Reasoning
1. The court avoided deciding the public-records exception
The trial court admitted the custody documents under Vermont Rule of Evidence 803(8)(A), which covers records of regularly conducted and recorded public-agency activities. Rule 803(8)(B), however, excludes specified categories, including certain police investigative reports, investigative reports offered by the government in its own case, governmental factual findings in criminal cases, and untrustworthy material.
Welch argued that the documents contained factual assertions offered to establish that the samples had not been altered. The Supreme Court deliberately left their classification unresolved. Consequently, this opinion does not hold that police submission forms or forensic custody reports generally qualify as admissible public records.
Instead, the court applied harmless-error analysis, consistent with V.R.Cr.P. 52(a), which requires disregarding errors that do not affect substantial rights. Five police officers and three laboratory employees described collection, sealing, securing, transfer, receipt, and testing procedures. That testimony established the required reasonable assurance independently of the four exhibits.
Welch’s strongest prejudice argument was that laboratory witnesses relied on the documents to testify about sample condition. The court answered that a document can refresh memory under V.R.E. 612 without becoming substantive evidence itself. Excluding the exhibits therefore would not necessarily remove the witnesses’ testimony or the DNA results.
Although the opinion recited all four harmless-error factors, its application principally emphasized the documents’ redundancy and the independent basis for authenticating the samples.
2. The allele-table challenge failed before an exception was needed
The analyst examined twenty-two genetic loci, generated DNA profiles, compared the relevant alleles, and entered his own results into the tables. He testified about both the underlying work and his conclusions.
The court observed that it was unclear whether the trial judge had admitted these tables under Rule 803(8) at all. More importantly, Welch had not demonstrated that they were hearsay in the first place. A challenge framed solely around the exclusions from a hearsay exception could not succeed without establishing the threshold need for that exception.
The court also cited V.R.E. 703, which permits an expert to base an opinion on facts or data perceived by or made known to the expert. That reference should not be read as making every document underlying expert testimony automatically admissible. The ruling was tied to this analyst’s personal testing, data entry, and testimony. The contrasting authority concerning another analyst’s findings underscores that limitation.
3. The instruction depended on the character of the evidence
Welch requested an instruction that consciousness-of-guilt evidence has relatively little weight and cannot alone support conviction. The court held that the calls did not require that instruction merely because the State characterized them, in part, as evidence of consciousness of guilt.
Unlike the alternate-name evidence in State v. Scales, these statements directly concerned the pending prosecution. They were made soon after arrest while Welch was detained on the charged offense. The jury did not need to speculate about which crime he had in mind.
The statements also bore on intent, a genuinely disputed issue given the sleep-related defense. References to the evidence, M.B.’s testimony, and maintaining an account could be evaluated alongside Welch’s explanations. Their weight remained for the jury.
Defense counsel remained free to argue—and did argue—that consciousness of guilt was insufficient to establish guilt. That opportunity was part of the trial context, not a general substitute for a legally required instruction. The decisive point was that these particular statements lacked the attenuated inferential chain that justified caution in the flight-related cases.