Keep Factually independent
Whether you agree or disagree with our analysis, these conversations matter for democracy. We don't take money from political groups - even a $5 donation helps us keep it that way.
What percentage of reported sexual assault cases are classified as false or unfounded, and how do definitions vary by agency?
Executive summary
Most rigorous studies and advocacy groups put confirmed false or “false report” rates of reported sexual assaults in the roughly 2–10% range, with several meta-analyses and law‑enforcement case‑classification studies clustering near 5–7% (examples: 5.2% meta‑analysis, 7% MAD study, 5.9% university series) [1] [2] [3]. However, many police “unfounded” or “baseless” designations in administrative data are higher in some agencies and can reflect investigatory practice, coding rules, or bias rather than independently verified false allegations [4] [5].
1. Why the headline numbers converge around 2–10% — and why that matters
A body of peer‑reviewed work and advocacy summaries report confirmed false‑report rates between about 2% and 10%; Claire Ferguson and John Malouff’s meta‑analysis and related reviews land near 5% as a central estimate, and multi‑site law‑enforcement research (the MAD project reported 7% coded false) and single‑institution studies show similar figures [1] [2] [3]. Journalists and national centers such as the National Sexual Violence Resource Center and major outlets repeat this 2–10% window because it reflects studies that apply explicit criteria for “false” (evidence showing no crime occurred) rather than looser internal labels [1] [2].
2. “Unfounded” vs. “false”: labels change the story
Police and reporting systems sometimes use “unfounded,” “baseless,” or “unsubstantiated” as closure codes that do not equate to an independently proven false allegation; the justice‑system literature warns these classifications can hide investigative failure or differing record‑keeping practices [4] [5]. ProPublica’s reporting found some agencies classifying more than 10% of rape reports as “unfounded” when internal records were examined, showing agency rates can diverge markedly from the academic confirmed‑false estimates [4].
3. How agencies define and decide “false” varies widely
Research shows the strict definition used by many academic studies requires affirmative evidence that the assault did not occur; police departments may apply informal judgement, evidentiary thresholds, or administrative convenience when coding cases “unfounded,” producing higher or inconsistent rates across jurisdictions [3] [6]. The FBI and federal reporting changes have broadened definitions of rape for counting incidents, but they don’t standardize how an investigation is closed as false versus unfounded at the local level [7] [5].
4. Institutional incentives and potential biases that affect classification
Observers and federal guidance note three drivers of “unfounding”: lack of training or misconduct, limited reporting categories in older systems, and discretionary officer judgement — all of which can inflate unfounded counts independent of actual false reporting [5] [8]. High unfounded rates have triggered scandals in many cities and prompted federal calls for better tracking; these patterns suggest some jurisdictional variation reflects institutional practice more than true variation in lying [4].
5. Settings where rates look different: campuses, military, and small samples
Some analyses of campus adjudication or particular agencies report much higher “unfounded” or “insufficient evidence” percentages (reports range widely; some advocacy or legal‑defense groups argue wrongful findings may be common), while other rigorous samples (meta‑analyses, forensic reviews) still show low confirmed‑false percentages [9] [10] [2]. For example, small or non‑representative studies can produce outlier rates (e.g., single‑department findings above 25% or historical small‑sample claims of very high false rates) that do not generalize without careful method review [10] [4].
6. What reporting gaps remain and how to interpret the evidence
Available sources show consensus that confirmed false reports are uncommon (centred ~5%), but they also repeatedly say “no one knows exactly” because police coding, investigative thoroughness, and dataset definitions vary; some experts caution that “unfounded” data in national systems are incomplete or manipulable [2] [4]. Current reporting does not provide a single, nationwide standardized percentage and notes that higher local unfounded rates may signal investigative problems rather than more lying [4] [5].
7. Practical takeaway for readers and journalists
When you see a single percentage cited, check which definition and data source it uses: “confirmed false” (evidence shows no crime) typically yields low estimates (~2–7% across rigorous studies), while agency “unfounded/baseless” labels can be much higher and reflect procedures or bias [1] [2] [4]. Reporters should ask whether a study used independent validation criteria, sampled multiple jurisdictions, and accounted for classification practices before treating any single number as definitive [3] [6].
Limitations: available sources do not provide a single authoritative, current national percentage that standardizes definitions across all agencies; many findings depend on differing methodologies and local practices [4] [2].