Answers are grounded in this paper and cite their sources.
Background Case-level evidence on homicide in schools comes almost entirely from the United States. We built a record for other countries from public sources and asked how complete it is, how its cases differ from US cases, and what rate comparison with the United States it can support. Methods We searched public sources, country by country, for homicides on K-12 school property from 2000 to August 7, 2026, and coded each case on the instrument used for a US database. The analysis covered the 193 World Health Organization member states other than the United States. Six checks tested completeness. We compared cases with 398 US cases from 2000 to 2025 and computed incidence per 100 million school-age person-years, keeping zero-case countries in every denominator, at thresholds of 1, 2, 4, and 8 or more homicide victims. Results We identified 335 incidents in 86 countries (732 deaths), of which 299 in 79 countries (654 deaths) occurred from 2000 to 2025. Countries holding 31.4% of school-age exposure contributed no cases. Retrospective searches found far more single-death incidents in recent years than earlier, while incidents killing 4 or more appeared at a steady rate. Sources for single-death incidents came mainly from national news (73.4%) and seldom from international outlets (14.1%); incidents killing 4 or more drew on both (51.5% and 60.6%). Perpetrators used knives in 51.2% of non-US incidents and firearms in 40.1%, against 15.8% and 80.2% in the United States. The US rate ratio was 38.7 (95% CI, 33.2–45.1) against all member states and 9.7 (7.8–12.0) against high-income countries. Excluding homicides outside the school building cut the high-income ratio to 3.8 (2.7–5.2). At 4 or more deaths, it was 3.3 (1.2–9.7); point estimates stayed between 2.7 and 4.4 under every case definition tested, but one interval included 1. Conclusions A media-based record of school homicide outside the United States appears to capture severe incidents far more completely than single deaths. It can show how incidents differ from US cases, most firmly by weapon, and it supports a rate comparison with the United States only as a range tied to comparison group, threshold, and case definition.
No takes yet. Share an insight, caveat, or question.
Densley et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: