Liberty Evidence Engine

Evidence · 2026-08-31

How much of U.S. homicide is FBI “active shooter”?

Mass-public-shooting headlines are not the homicide problem. FBI active-shooter reports name a real, narrow category. Put those killed counts next to the same national murder estimate LEE already uses for urban concentration, and the share is a fraction of one percent. Download the series. Not a gun-law causal model.

Download series CSV Adjust the years Methods

Interactive explorer · FBI active shooter vs homicide

Adjust the years. The 2023 share stays a scale fact.

Loading FBI series…

Window

—

Incidents in window

—

Killed in window

—

2023 share of murder estimate

—

FBI active shooter ≠ Gun Violence Archive “4+ shot” ≠ media “mass shooting.” 2023 denominator is the same FBI murder estimate already locked on the urban homicide page (18253). Recomputed in your browser from the downloadable CSV — not a gun-law causal model.

Year series

Static reference table. Use the explorer above to try other settings.

year incidents killed wounded mass_killings fbi_murder_estimate share_of_murder_pct
2018 30 86 139
2019 30 103 155 13
2020 40 38 126 5
2021 61 103 140 12
2022 50 100 213 13
2023 48 105 139 15 18253 0.575

Methods & limits

Question. When policy talk treats “mass shooters” and their gun purchases as the face of American homicide, how large is the FBI’s own active-shooter killed count relative to estimated U.S. murder?

  • This is a scale exhibit, not a threat-assessment model and not a re-run of any state’s firearm-purchase (DROS) database. We do not have California dealer-record microdata and we do not train a rival classifier.
  • Numerator is people killed in FBI-designated active shooter incidents. The FBI’s working definition is one or more individuals actively engaged in killing or attempting to kill people in a populated area. The 2023 report is explicit that this is not all gun crime: it excludes, among other things, gang-and-drug related shootings and many residential or domestic disputes.
  • FBI “mass killing” on this page is the federal three-or-more-killed-in-a-single-incident definition used in those reports — not Gun Violence Archive “four or more shot,” not a media “mass shooting” list, and not “every homicide with a firearm.” Definition chips are prose, not a second database we did not ingest.
  • 2019–2023 incidents, killed, wounded, casualties, and mass-killing counts are taken from the FBI/ALERRT PDF Active Shooter Incidents in the United States, 2023 (June 2024). The 2019–2023 killed table in that PDF is: 2019 103; 2020 38; 2021 103; 2022 100; 2023 105. Incidents: 30, 40, 61, 50, 48. The same PDF states 48 incidents and 244 casualties (105 killed, 139 wounded) for 2023.
  • 2018 incidents (30) and killed/wounded (86 / 139) come from the FBI/ALERRT PDF Active Shooter Incidents in the United States in 2022 (April 2023). The 2023 PDF says prior published incident data was not updated in that later report, so 2018 is labeled from the 2022 PDF only.
  • 2023 denominator is 18,253 estimated U.S. murder and nonnegligent manslaughter — the same FBI Crime in the Nation 2023 figure already locked on LEE’s urban homicide concentration page. Share = 105 / 18,253 = 0.575%. Later restatements of the 2023 murder total exist in secondary compilations; this page does not silently switch denominators so the two LEE pages still compose.
  • Share of murder is computed only for 2023, the year we have that locked denominator. Other years show FBI active-shooter counts without a fake year-by-year murder series from Statista or news roundups.
  • Interactive explorer recomputes in the browser from the downloadable CSV: year window, killed/incidents in the window, 2023 share callout, shareable URL (?start=&end=).
  • This page does not estimate the effect of carry laws or ownership elasticities. Those questions are out of scope here.
Limitations. Active-shooter “killed” is not an accounting identity inside UCR murder: the FBI category includes some law-enforcement deaths in those incidents, and UCR murder is a different bookkeeping system. A 0.58% share means this category is small relative to homicide, not that the incidents are unimportant and not that every mass killing is inside the FBI active-shooter file. 2014–2017 annual killed totals are not on this page because we did not extract a clean year table from those earlier PDFs. City and neighborhood concentration of ordinary homicide is a separate LEE page. This page does not test whether mass shooters buy guns on an unusual timetable.

Sources.