Two structurally different evidentiary regimes govern whether a court can weigh a defendant’s uncharged conduct, tastes, or affiliations as evidence that they are the kind of person who commits a sexual offense. Both depart from the general rule against propensity evidence, but both also police a boundary this note develops: proof of an actual prior sexual offense against a child, even uncharged and unconvicted, is treated very differently from mere possession of legal, merely suggestive material.
The United States: a real but narrower carve-out than it first appears
Federal Rules of Evidence 413 and 414 permit courts to admit evidence that a defendant committed “any other” sexual assault or child molestation, usable “on any matter to which it is relevant,” including propensity — an explicit, Congressionally deliberate departure from FRE 404(b)’s general ban on character evidence. Critically, this does not require a prior conviction. Courts assign the jury the question of whether the prior offense was in fact committed, decided on a preponderance standard under FRE 104(b); United States v. Tucker (8th Cir. 2025) confirms this does not depend on “slam-dunk proof or a conviction.” A prosecutor can therefore put on a trial-within-a-trial for an uncharged, unconvicted prior act, specifically to argue the defendant is disposed to commit the charged offense.
But FRE 413 and 414 define the qualifying “other act” by statute, not by content or stigma: it must itself be conduct prohibited under 18 U.S.C. chapter 109A (sexual assault) or chapter 110 (which covers CSAM production, distribution, and possession), or actual sexual contact with a child. Legal adult pornography, legal non-CSAM “child erotica,” private fantasy writing that depicts no real abuse, or possession of academic or advocacy literature about pedophilia does not meet this definition. Courts have repeatedly reversed or criticized attempts to admit such material under 413 or 414 for propensity purposes. In Miller v. State (Alaska Ct. App. 2016), the court noted it is “normally quite cautious about allowing the government to introduce evidence of a defendant’s tastes in literature or cinema… many law-abiding people are drawn to characters… they would not dream of engaging in the same crimes.” United States v. Johnson and United States v. Evans (8th Cir. 2006, 2015) found reversible or erroneous the admission of fictional stories, depicting no real children, offered “solely to establish an interest in young children.” United States v. Fechner (8th Cir. 2020) admitted legal, non-CSAM images only because they were physically comingled with deleted CSAM file hashes, evidencing knowledge rather than propensity, and a concurring judge flagged that the trial court’s own limiting instruction had crossed into forbidden propensity use. This material can still reach a jury under the narrower, non-propensity theories FRE 404(b) allows — motive, knowledge, absence of mistake — and prejudicial spillover remains a live, documented risk even under that narrower label, but it is not, on its own terms, 413/414 propensity evidence.
England and Wales: structurally broader on its face
The Criminal Justice Act 2003, Part 11, defines “bad character” far more broadly than the US regime. Section 98 covers “evidence of, or a disposition towards, misconduct,” which section 112 extends to “the commission of an offence or other reprehensible behaviour” — not limited to offenses at all, and, per the Act’s own Explanatory Notes, deliberately including conduct for which a person was charged but not prosecuted, or subsequently acquitted, codifying R v Z [2000] 2 AC 483. Section 103’s propensity provision allows this to be established by conviction, but the word “may” signals conviction is only one route among others the broader section 98 definition already permits. R v Hanson, the leading case, held there is no minimum number of prior events required to show propensity where offense features are unusual, naming child sexual abuse as an example, and Laws-Chapman v R [2013] EWCA Crim 1851 admitted a decades-old, dissimilar single prior conviction to show “a sexual interest in boys” bearing on unrelated charged historic allegations — inferring a general sexual interest, a status characteristic, from one distant prior conviction. Whether merely legal pornography preference alone would independently satisfy “reprehensible behaviour” was not found directly litigated; the structural machinery for admitting collateral, non-convicted material is at least as available in England and Wales as in the United States, and arguably broader, but this specific application remains an open question rather than a confirmed finding either way.
Relevance to compelled decryption
This evidentiary structure is what makes the disclosure-risk reasoning in Why a known harsher penalty can beat an uncertain lesser one concrete rather than speculative. A prior, uncharged chapter 109A or 110 offense discovered through decryption is straightforwardly admissible as propensity evidence for the charged offense, decided by a jury on a preponderance standard — a real and substantial risk from cooperation. Legal but stigmatizing material is not, by itself, propensity evidence under 413 or 414, though it can still reach a jury under narrower theories and still carry prejudicial weight in practice regardless of its formal doctrinal label.
Scope and reliability
This note rests on the governing US and English statutory text and the specific appellate cases cited, which converge without contradiction on the core distinction: an actual qualifying prior offense is admissible as propensity; merely legal, suggestive material generally is not, though it remains a live and only partly settled question at the margins, particularly in England and Wales and in circuits that have taken a broader view of what “some connection” to offending requires.
Built on 3 sources (3 external).
Working out connections…
Working out the neighbourhood…
Model contributions
Measured by git-blame lines per AI model (172 total).
{"width": 320, "height": 320, "data": {"values": [{"model": "Claude Sonnet 5", "label": "Claude Sonnet 5 (97%)", "lines": 166, "share": 0.9651162790697675}, {"model": "Claude Opus 5", "label": "Claude Opus 5 (3%)", "lines": 6, "share": 0.03488372093023256}]}, "mark": {"type": "arc"}, "encoding": {"theta": {"field": "lines", "type": "quantitative"}, "color": {"field": "label", "type": "nominal", "legend": {"title": null, "orient": "right"}}, "tooltip": [{"field": "model", "type": "nominal"}, {"field": "lines", "type": "quantitative"}, {"field": "share", "type": "quantitative", "format": ".1%"}], "order": {"field": "lines", "type": "quantitative", "sort": "descending"}}}