Background
Israel Rosa Figueroa was convicted by a Superior Court jury on thirteen indictments charging him with the sexual abuse of two minor children, both relatives. The jury returned six counts of aggravated rape of a child, two counts of rape, and four counts of indecent assault and battery against victim one, and one count of indecent assault and battery against victim two. The Commonwealth had obtained separate sets of indictments in 2018 and 2019. When the 2019 indictments — naming two additional relatives, JG and KG — were severed before trial, the trial judge advised the prosecutor that the conduct underlying those charges could be offered as prior bad act evidence at the trial on the 2018 indictments. The judge allowed that motion over the defendant’s objection.
The resulting trial featured testimony from all four witnesses. Victim one, abused from approximately age nine to fifteen, testified about both the charged conduct and substantial uncharged acts. Victim two, who was about six years old during the abuse, described the charged offense and additional uncharged conduct. JG testified that the defendant abused her every day from age eight to sixteen, including penetration, photography of intimate acts, and a request that she move to Puerto Rico with him so he could impregnate her. KG testified to daily abuse from ages four to seven. Testimony about uncharged or prior bad acts consumed forty-two transcript pages — with twenty-nine pages devoted solely to JG and KG — while testimony from the named victims about the charged offenses covered only twenty-six pages. The prosecutor’s opening announced that the defendant “was molesting the female children in this family,” and her closing urged conviction because he “preyed upon the children in this family.” The judge’s limiting instructions told the jury it could consider the prior bad act evidence for a lengthy list of purposes: “motive, opportunity, state of mind, intent, preparation, plan, pattern of conduct, relationship between the defendant and a victim, knowledge, identity, absence of mistake, or accident.”
The Court’s Holding
Judgments vacated; verdicts set aside. Writing for the panel, Judge Smyth held that the trial judge abused her discretion by failing to prevent the volume and detail of prior bad act evidence from overwhelming the evidence of the charged conduct. Following the Supreme Judicial Court’s reasoning in Commonwealth v. Dwyer, 448 Mass. 122 (2006), the Appeals Court found that a trial judge has an independent obligation to intervene — even without defense objection — when accumulating prior bad act testimony threatens to divert the jury from the charged offenses. That obligation went unmet here.
The panel identified several reinforcing failures. First, the numerical disproportion was stark: forty-two transcript pages of prior bad act testimony versus twenty-six pages on the charged conduct, with the two prior bad act witnesses alone accounting for twenty-nine of those forty-two pages. Second, the prosecution’s opening and closing amplified the prior acts into a propensity narrative — the defendant as serial predator of female relatives — rather than confining the jury’s attention to the specific charged crimes. Third, the limiting instruction was a “laundry list” of permissible purposes rather than a narrow, specific instruction tied to any particular piece of evidence — exactly the practice the SJC condemned in Commonwealth v. Samia, 492 Mass. 135, 148 n.8 (2023): “Proffering a laundry list of nonpropensity purposes is not helpful, nor is it proper.” Even if the instructions had been narrow and forceful, the court held, the sheer volume of the prior bad act testimony in a child sexual abuse case was so great that curative instructions were insufficient: “asking the jury to disregard it may be tantamount to asking the jury to ignore that an elephant has walked through the jury box.” Because there was no corroborating forensic evidence and the case turned entirely on witness credibility, the defendant was entitled to a new trial should the Commonwealth choose to pursue one.
Key Takeaways
- Prior bad act evidence in child sexual abuse prosecutions must be proportionate to the evidence of the charged conduct. When testimony about uncharged acts substantially exceeds testimony about charged offenses — here, roughly 42 to 26 transcript pages — the trial judge has an independent duty to intervene and limit the volume, even without a timely defense objection.
- Limiting instructions, however well-crafted, cannot cure the prejudice caused by an overwhelming quantity of prior bad act evidence in cases where the verdict hinges on witness credibility without corroborating forensic proof. The “elephant in the jury box” analogy reflects the practical limits of any curative instruction.
- Prosecutors and trial judges must avoid “laundry list” justifications for prior bad act evidence. Under Samia, counsel must articulate the specific, narrow nonpropensity purpose for each piece of such evidence, and the court should limit its jury instruction to that precise purpose — not a menu of possibilities.
- Severance does not immunize severed conduct from prejudice analysis. When severed indictments are re-offered as prior bad act evidence, the trial judge must rigorously manage both the admissibility and the volume of that testimony, particularly where the uncharged conduct mirrors the charged conduct in type and frequency.
Why It Matters
Commonwealth v. Figueroa is a significant decision for both the criminal defense bar and prosecutors handling child sexual abuse cases in Massachusetts. It confirms that the evidentiary gateway for prior bad acts is not a one-way door: once prior bad act witnesses are allowed at all, the trial judge retains an ongoing supervisory obligation to prevent testimony from overwhelming the case. When that obligation goes unfulfilled — and when the prosecution’s own framing invites the propensity inference — the resulting convictions cannot stand even if each individual piece of prior bad act evidence was independently admissible.
For the Commonwealth, the decision highlights the risk of a charging strategy that severs indictments and then presents the severed conduct as prior bad acts. While such testimony may be admissible in principle, its quantity and courtroom emphasis require active judicial management. Defense practitioners should expect that post-Figueroa challenges to prior bad act evidence will focus not only on threshold admissibility but on whether the trial record reflects the proportionality and specificity that Massachusetts law demands.