The FBI’s Active Shooter Reports Exclude Self-Defense by Rule, and Its Own Founding Study Says So
A new CPRC analysis puts armed-citizen stops at 36.6 percent against the 3.9 percent drawn from FBI reports. We pulled the FBI’s founding study: the gap starts with a printed exclusion rule.
Armed citizens stop active shooters far more often than the number you’ve been told, and the reason isn’t that the FBI can’t find the cases. It’s that the FBI’s active shooter reports are compiled under a written rule that excludes incidents research establishes were self-defense.
Simply put: the count most people quote to prove armed civilians rarely matter was built, on paper, to leave out the cases where they did.
On September 9 the Crime Prevention Research Center published an analysis by John R. Lott of the FBI’s reports for 2014 through 2025. CPRC counts 408 incidents in the bureau’s reports, 16 stopped by an armed citizen, or 3.9 percent. We’ll show you where a number like that ends up.
What these reports are: The FBI contracts the research to the Advanced Law Enforcement Rapid Response Training Center at Texas State University, then reviews and refines it. The series rests on a founding study, A Study of Active Shooter Incidents in the United States Between 2000 and 2013, and we pulled it ourselves.
The rule, in the bureau’s own words: The study defines an active shooter as “an individual actively engaged in killing or attempting to kill people in a confined and populated area.” Then it lists what it threw out. An event was excluded if it primarily involved “Conflicts arising from self-defense; Gang violence; Contained residential or domestic disputes; Controlled barricade/hostage situations; Crossfire as a byproduct of another ongoing criminal act; or Drug violence.”
Let’s be clear here – a study that removes self-defense cases before it starts counting can’t then tell you how often citizens defend themselves against active shooters, any more than a crash study that drops every collision where the airbag deployed can tell you whether airbags work.
The study never claims otherwise. It calls itself “not intended to explore all facets of active shooter incidents, but rather is intended to provide a baseline.” Even so, 5 of its 160 incidents (3.1 percent) ended after “armed individuals who were not law enforcement personnel exchanged gunfire with the shooters,” so the count isn’t zero, only thin, and the rule is the best explanation for why.
The rule has softened on paper since, and it changed nothing. The 2021 and 2022 reports say an incident “was excluded if research established that it was the result of: Self-defense”; the 2024 report says one “may be excluded if it involves: self-defense.” CPRC says the 2024 report still includes no self-defense cases, though CPRC found 18 that met the FBI’s own definition, which in its words “surely creates the impression that such cases don’t exist.”
What CPRC found: Applying the FBI’s definition without dropping self-defense, CPRC counts 615 incidents over the same years, 225 stopped by an armed civilian, or 36.6 percent. Where carrying was legal, citizens stopped 199 of 376 attacks, 52.9 percent, and within that carry-legal subset the 2025 figure alone was 58.3 percent (21 of 36); CPRC’s all-locations 2025 rate is 42.6 percent. CPRC published the spreadsheet and a page linking the news story behind every case.
Where the number ends up: Here’s what that figure does once it leaves the report. On July 17, 2022, Jonathan Sapirman, 20, opened fire in the Greenwood Park Mall food court in Greenwood, Indiana. He killed Pedro Pineda, 56, Rosa Mirian Rivera de Pineda, 37, and Victor Gomez, 30, and wounded a 22-year-old woman and a 12-year-old girl.
Elisjsha Dicken, 22, of Seymour, Indiana, lawfully carrying a Glock 19, shot him. Greenwood Police Chief Jim Ison said surveillance timestamps put 15 seconds between the first shot and Dicken ending it. Ison put the distance at 40 to 50 yards; Dicken fired 10 rounds, and the autopsy found 8 gunshot wounds on the attacker.
Indiana’s permitless carry law, House Enrolled Act 1296, had taken effect July 1, 2022, sixteen days before the attack, meaning Indiana adults could lawfully carry there without a permit; the law was barely two weeks old when it was tested. The Johnson County prosecutor declined to charge Dicken, finding the shooting justified.
The next day, Reuters ran a story headlined “Armed bystander credited with preventing more deaths in Indiana shooting.” It still told readers: “It is rare for a bystander to stop an ‘active shooter’ attack in the United States, according to an FBI report in May showing that only two out of 61 such attacks last year ended when ‘citizens engaged the shooter.'” One day after an armed citizen ended an attack in 15 seconds, a global wire service told its audience this is rare, citing a report that excludes self-defense by rule.
Therein lies the whole problem: nobody at Reuters was lying; the number was filtered before they saw even it, and the filtered number is now what everyone knows.
Dearborn, and the word on the file: On July 3, 2026, Cameron Watkins, 19, of Detroit, shot and killed Keonte Seaborn, 19, inside Fairlane Town Center in Dearborn, Michigan, then kept firing and wounded a third person, whose injuries weren’t life threatening. Martinez Long, 19, of Oak Park, Michigan, shot and killed Watkins. The Wayne County Prosecutor’s Office found Long acted in self-defense and in defense of others and filed no homicide charge.
Long was still charged with carrying a concealed weapon without a permit, a felony, and on September 8, three days ago, he was sentenced to two years of probation under a youthful-offender provision that can wipe his record. Now look at the finding the prosecutor attached to the shooting itself: self-defense. That’s the first item on the FBI’s exclusion list, which means the case that should absolutely be counted in the statistics is the one the rule is written to drop.
The fair objection: You should be skeptical of Lott. He’s an advocate, CPRC’s 615-case universe is broader than the FBI’s 408, and it’s built largely from news searches. CPRC itself concedes the recency bias: recent cases are easier to find, which is part of why its rate climbs from 17.4 percent in 2014 to 42.6 percent in 2025, so the 36.6 figure has real give in it.
None of that rescues the 3.9, because the exclusion rule doesn’t come from Lott. It’s printed in the FBI’s own study, in the FBI’s own words, and you can read it without trusting a single number CPRC produced.
Springfield Echelon Alpha 4.0C (9mm, optic-ready)
Optic-ready, 15+1, and a full 4-inch barrel on a compact frame, which is the combination that makes hits past conversational distance realistic. The Alpha trim strips the Echelon's extras and still lands under $500 at Palmetto State Armory.
What to take from this: Armed citizens stop active shooters, and the FBI’s own founding study records it happening. What the FBI’s reports can’t tell you is how often, because they remove self-defense cases before the count begins, and every number built on them, the 3.9 percent, the “two out of 61,” inherits that removal. A count that excludes the outcome you’re asking about isn’t evidence about that outcome.
It’s the same lesson we drew from the CDC’s defensive gun use estimates: read how a federal number was defined before you repeat it. Then watch the FBI’s next annual report for Dearborn. If it’s there, the rule has changed; if it isn’t, you’ll know exactly why.
Sources: the Crime Prevention Research Center analysis; U.S. Department of Justice and Federal Bureau of Investigation, A Study of Active Shooter Incidents in the United States Between 2000 and 2013; Greenwood Police Department briefings as reported by WRTV, WTHR, WISH-TV and WFYI; the Wayne County Prosecutor’s Office as reported by the Detroit News, WXYZ, CBS News Detroit and Fox 2 Detroit. Featured photo: Carol M. Highsmith, Library of Congress, public domain.
Justin Trump is the managing editor and owner of CAT Outdoors. The son of a Vietnam veteran, he’s a Certified Glock Armorer, an avid gun enthusiast and 2A advocate. He holds two firearm patents for the CAT M4 and Talon tools. When not managing CAT Outdoors, he enjoys spending time with his family and friends, rooting for Michigan sports teams, and serving his church.