FBI Says Armed Citizens Stop 4% of Shooters; Lott Says 37%
John Lott's Crime Prevention Research Center says armed citizens stopped 36.6% of active-shooter incidents between 2014 and 2025. The FBI's count for the same window credits civilians with 3.9%. That is roughly a tenfold difference, and it is not a rounding error. Lott challenged the federal numbers in a October 23 National Review article, and the CPRC's latest research backs him up. The FBI logged 408 incidents across those years. The CPRC logged 615. The real dispute is not about how many shootings happened. It is about what counts as a shooting that was stopped.
Key Details
The numbers break down this way:
- FBI count, 2014 to 2025: 408 active-shooter incidents nationwide.
- FBI credit to armed citizens: 16 attacks stopped, or 3.9% of the total.
- CPRC count, same period: 615 active-shooter attacks identified.
- CPRC credit to armed citizens: 225 attacks stopped, or 36.6%.
- Most recent five years in the CPRC data: the stop rate rose to 42.5%.
- Preshooting interventions: the CPRC found 35 more incidents where armed citizens stopped an attacker before any shots fired. The CPRC left these out of its rate because no shooting occurred.
- Missed by the FBI: the CPRC says the bureau missed 51 incidents that likely would have become mass public shootings, plus 156 other active-shooting incidents.
Lott also argues the FBI misclassified several cases. The bigger problem, in his telling, is omission. The FBI's method drops incidents that never grew large enough to qualify. Those are often the cases where an armed citizen ended the threat early. The CPRC's own examples include recent high-profile episodes in Twin Falls, Idaho, and Bend, Oregon, where armed civilians engaged attackers and ended the threat. Those cases drew national coverage because of the body count. Thousands of smaller incidents never did.
Background
John Lott built his reputation on the claim that right-to-carry laws reduce violent crime. His 1998 book More Guns, Less Crime drew on county-level data and became a central text in the debate. Critics have challenged his methods for decades, and the argument has never fully settled. Lott's core complaint is different. He has attacked federal and academic research regardless of which party controls the government, arguing that official datasets undercount defensive gun use.
The FBI publishes active-shooter reports that define an incident and record outcomes, including whether an armed civilian intervened. Those reports feed media coverage and policy debates. Advocates on both sides cite them. The NRA's line that the only thing that stops a bad guy with a gun is a good guy with a gun dates to the post-Newtown debate in 2012, and it has drawn criticism ever since. In New York State Rifle & Pistol Association v. Bruen in 2022, the Supreme Court set a text-and-history test for gun regulations. That test made historical and empirical claims about armed self-defense more relevant to litigation, not just op-eds.
Why It Matters for Gun Owners
Every carrier should care about this gap, because policymakers use the FBI number to argue that civilian intervention is rare and risky. If the federal figure is the baseline, a legislator can claim that concealed carry offers little public safety benefit. A higher stop rate changes that conversation. It also helps defend the idea that lawful carriers are part of the solution, not a separate problem.
The impact reaches beyond national headlines. Anyone carrying in a state with permitless carry, shall-issue permitting, or a recognized reciprocity agreement is affected by how the data gets framed. Legislatures in states that restrict carry will cite the FBI figure in hearings. Gun owners in those states should know the CPRC data exists and where it comes from. Bring the source, not a talking point.
The practical message is simple. Carriers need training that covers threat recognition, movement, and when not to draw. The same research that counts stops also counts shots fired by people who should not have fired. Lott's point about preshooting interventions shows that many good outcomes involve no shots at all. Deterrence and visible readiness matter, but so does staying out of the way of police once the shooting starts.
DownRange Analysis
The FBI's number is a classification problem before it is a statistics problem. If the bureau counts only incidents that reach a certain body count, then its own filter guarantees that armed civilians look rare. Lott's critique is the right one to press. Defensive gun use data should count outcomes, not just casualties.
Under Bruen, government claims about public safety still need historical and empirical support when they restrict arms. A federal dataset that understates civilian intervention is a weak foundation for that kind of argument. Expect gun-control litigants to challenge the CPRC methodology. Expect Lott's critics to argue that 35 preshooting interventions and a broader definition inflate the count. Both sides will fight over definitions. Gun owners should insist on transparent criteria.
Right now, do three things. Read the CPRC research and the National Review piece before you quote either. Keep your training current, because a stop depends on skill, not just a holster. And if you carry in a state with active legislation, track which lawmakers cite the FBI's 3.9% figure. Ask them why they trust a count that drops incidents before they are recorded.
