Abstract
Administrative registries have become central instruments of education policy against sexual violence in schools, yet their figures reflect both the violence that occurs and the state’s capacity to detect it. This study analyses the complete open data series of Ecuador’s Violence Registration System (REDEVI) for 2014–2025, comprising 33,020 cases and 34,229 recorded victims, linked to national enrolment records. The analysis combines enrolment-based rates, a segmented quasi-Poisson regression with sensitivity analyses, and comparisons by perpetrator group, date definition, educational level and sex. The recorded rate rose from 11.9 to 109.6 victims per 100,000 students between 2016 and 2018, as the registry was institutionalised, but the annual series cannot distinguish this discontinuity statistically from the start-up trend. Four strands of evidence are consistent with surfacing, namely, the Ministry’s own attribution, the retrospective entry of cases in 2018, a shift towards externally perpetrated violence detected by schools and a 56% contraction during school closures, deepest for within-system cases. Recorded victimisation peaked in upper basic education and was 8.8 times higher among students recorded as female. The state’s capacity to count school-related sexual violence depends on protocols, in-person schooling and data infrastructure, which makes registry governance a component of education policy.