State v. Duran, 343 P.3d 207

Defendant Jesse Duran was convicted of first-degree criminal sexual penetration of a minor based primarily on the victim’s testimony that Duran digitally penetrated her when she was in elementary school; the victim testified years later and had delayed reporting the abuse.
The victim told family members at various times; formal law enforcement investigation and a S.A.F.E. House forensic interview occurred in 2005; trial occurred after charges filed in 2006.
Denise Clement, a S.A.F.E. House forensic interviewer, testified about her interview of the victim and her general experience conducting roughly 1,400–1,600 child forensic interviews from 2002–2008.
On voir dire and before the jury, the prosecutor elicited from Clement that, based on her experience and recollection of internal data, a majority of children she interviewed delayed disclosure of sexual abuse.
The defense objected that this subject required expert testimony; the district court allowed Clement to offer the frequency/delayed-disclosure testimony as lay opinion under Rule 11-701.
The Court of Appeals held the testimony was expert (specialized) in nature, its admission as lay testimony was error, the error was not harmless because it bore on the central credibility issue, and therefore reversed and remanded for a new trial.

Issue	Plaintiff's Argument	Defendant's Argument	Held
Admissibility of testimony about frequency of delayed disclosure by child abuse victims	State: Clement’s statement was personal observation and within lay witness Rule 11-701 because based on her experience interviewing children	Duran: Frequency/general behavior evidence requires expert qualification under Rule 11-702; not proper lay opinion	Court: Admission as lay opinion was erroneous — frequency/delayed-disclosure testimony is specialized and must be offered by a qualified expert
Prejudice/harmless-error from admission of delayed-disclosure testimony	State: Any error was harmless because other evidence (victim testimony, mother’s testimony) supported conviction	Duran: Testimony was highly probative on credibility and likely affected verdict	Court: Error was not harmless — testimony was from State’s witness, introduced a non-cumulative new fact bearing on central credibility issue, creating a reasonable probability it affected the verdict

State v. Tollardo, 275 P.3d 110 (N.M. 2012) (non-constitutional harmless-error standard and factors to assess evidentiary error)
State v. Serna, 305 P.3d 936 (N.M. 2013) (evaluate circumstances around evidentiary error; source and emphasis factors)
State v. Leyba, 289 P.3d 1215 (N.M. 2012) (use of non-objected evidence to assess role of erroneously admitted evidence)
State v. Marrington, 73 P.3d 911 (Or. Ct. App. 2003) (erroneously admitted delayed-reporting testimony harmful where case was credibility contest)
State v. Gonzalez, 834 A.2d 354 (N.H. 2003) (lay testimony about frequency of delayed disclosure or recantation requires expert testimony)
State v. Dowling, 257 P.3d 930 (N.M. 2011) (double jeopardy does not bar retrial where sufficient evidence supported conviction)