Data: For convenience, a PDF of the Excel file is also available here. FBI Active Shooting reports are available here (2000-2013), here (2014-2015), here (2016-2017), here (2018), here (2019), here (2020), and here (2021).
Introduction
The shooting that killed three people and injured another at a Greenwood, Indiana, mall on July 17 drew broad national attention because of how it ended – when 22-year-old Elisjsha Dicken, carrying a licensed handgun, fatally shot the attacker.
While Dicken was praised for his courage and skill – squeezing off his first shot 15 seconds after the attack began, from a distance of 40 yards – much of the immediate news coverage drew from FBI-approved statistics to assert that armed citizens almost never stop such attackers: “Rare in US for an active shooter to be stopped by bystander” (Associated Press); “Rampage in Indiana a rare instance of armed civilian ending mass shooting” (Washington Post); and “After Indiana mall shooting, one hero but no lasting solution to gun violence” (New York Times).
Evidence compiled by the Crime Prevention Research Center shows that the sources the media relied on undercounted the number of instances in which armed citizens have thwarted such attacks by an order of more than ten, saving untold numbers of lives. Of course, law-abiding citizens stopping these attacks are not rare. What is rare is national news coverage of those incidents. Although those many news stories about the Greenwood shooting also suggested that the defensive use of guns might endanger others, there is no evidence that these acts have harmed innocent victims.
The FBI reports that armed citizens only stopped 11 of the 252 active shooter incidents it identified for the period 2014-2021. The FBI defines active shooter incidents as those in which an individual actively kills or attempts to kill people in a populated, public area. But it does not include those it deems related to other criminal activity, such as a robbery or fighting over drug turf.
An analysis by my organization identified a total of 360 active shooter incidents during that period and found that an armed citizen stopped 124. A previous report looked at only instances when armed civilians stopped what likely would have been mass public shootings. There were another 24 cases that we didn’t include where armed civilians stopped armed attacks, but the suspect didn’t fire his gun. Those cases are excluded from our calculations, though it could be argued that a civilian also stopped what likely could have been an active shooting event.
The FBI reported that armed citizens thwarted 4.4% of active shooter incidents, while the CPRC found 34.4%.
Two factors explain this discrepancy – one, misclassified shootings; and two, overlooked incidents. Regarding the former, the CPRC determined that the FBI reports had misclassified five shootings: In two incidents, the Bureau notes in its detailed write-up that citizens possessing valid firearms permits confronted the shooters and caused them to flee the scene. However, the FBI did not list these cases as being stopped by armed citizens because police later apprehended the attackers. In two other incidents, the FBI misidentified armed civilians as armed security personnel. Finally, the FBI failed to mention citizen engagement in one incident.
For example, the Bureau’s report about the Dec. 29, 2019 attack on the West Freeway Church of Christ in White Settlement, Texas, that left two men dead does not list this as an incident of “civic engagement.” Instead, the FBI lists this attack as being stopped by a security guard. A parishioner, who had volunteered to provide security during worship, fatally shot the perpetrator. That man, Jack Wilson, told Dr. John Lott that he was not a security professional. He said that 19 to 20 members of the congregation were armed that day, and they didn’t even keep track of who was carrying a concealed weapon.
As for the second factor — overlooked cases — the FBI, more significantly, missed 25 incidents identified by CPRC where what would likely have been a mass public shooting was thwarted by armed civilians. There were another 83 active shooting incidents that they missed.
There is no reason to think that the news media covers all the cases where civilians stopped attacks. And the farther back in time we go, the more cases we are likely to miss. The next table illustrates this bias. Using the 2014 to 2021 data clearly shows that 49.1% of active shooting attacks were stopped in 2021, 45.1% in 2020, and a declining percentage the farther back in time that we go. That pattern is consistent with us having a more difficult time finding cases that occurred farther in the past.
There is yet another reason that these corrected percentages are biased downward as they ignore that about half of these attacks occur where guns are banned, so law-abiding citizens who obey those rules wouldn’t have a chance to stop them. A copy of our Excel file and links for all the cases so people can check them are available here.
The FBI’s active shooting reports do not mention whether the attacks occur in gun-free zones. “The issue is that when places are posted as gun-free zones, law-abiding citizens obey those rules and would be unable to stop the attacks in those areas,” notes Carl Moody, a professor at William & Mary and the CPRC’s research director.
Surveys show that criminologists and economists had the same top four preferred policies for stopping mass public shootings. On a 1 to 10 scale where 1 was the least effective policy and 10 the most, American criminologists rated the following policies most highly: Allow K-12 teachers to carry concealed handguns (6.0), allow military personnel to carry on military bases (5.6), encourage the elimination of gun-free zones (5.3) and relax federal regulations that pressure companies to create gun-free zones (5.0). The top four policies for economists were the same, but in a different order: encourage the elimination of gun-free zones (7.9), relax federal regulations that pressure companies to create gun-free zones (7.8), allow K-12 teachers to carry concealed handguns (7.7), and allow military personnel to carry on military bases (7.7).
Between 2014 and 2021, citizens stopped 104 out of 204 potential or actual mass shootings where we could identify that guns were allowed in the area. So 51% of attacks were stopped by people legally carrying concealed handguns. Again, the most recent data is most accurate, and for 2021, 58% of the attacks were stopped in areas where people were clearly allowed to carry.
The numbers indicate If we didn’t have gun-free zones, we would have more people stopping these attacks.
Finally, even these numbers underestimate the usefulness of legally carried concealed handguns in stopping mass public shootings because many of these active shooting incidents involve only one person being targeted. For example, suppose one person is targeted and only one person may be present. In that case, there is relatively little opportunity for people to stop attacks compared to a mass public shooting where many potential victims are present.
The general public seems to agree. An early July survey by the Trafalgar Group showed that a plurality of American general election voters believe that armed citizens are the most effective element in protecting you and your family in the case of a mass shooting. First on the list was “armed citizens” at 42%, followed by “local police” (25%) and “federal agents” (10%). [“None of the above” was the answer chosen by 23% of respondents.] A survey by YouGov in May – before the Uvalde, Texas, attack – found that by a margin of 51% to 37% American adults supported letting schoolteachers and administrations carry concealed handguns.
Do concealed handgun permit holders who stop these attacks pose a danger to others?
News outlets often raise concerns that allowing concealed handgun carry will result in innocent bystanders being shot or in police accidentally shooting permit holders. White’s AP dispatch on the Greenwood shooting quoted Adam Lankford, identified as “a criminal justice expert at the University of Alabama,” who stated: “While it’s certainly a good thing in this mall shooting that someone was able to stop it before it went any further, let’s not think we can substitute that outcome in all past and future incidents. If everyone’s carrying a firearm, the risk that something bad happens just gets much larger.”
Professor Moody, who studies mass public shootings, notes that such warnings are misleading:
“The media and gun control advocates always seem concerned with the worst possible outcomes when firearms are involved. We know that armed citizens do, in fact, stop active shooters. And while there’s a possibility of a bystander getting hurt, the data put together by the CPRC show that an armed citizen has yet to accidentally shoot an innocent bystander. We also know that the police have accidentally shot the hero citizen just once. That was in Colorado on June 21, 2021. That’s not something that would normally happen, because the police usually arrive long after the incident is resolved.“
All the experts interviewed by the Washington Post and New York Times argue that stopping these attacks should be left to the police. “I think you might get more individuals carrying, sort of primed for something to happen, which is particularly dangerous … in reality that’s the job of the police,” Indiana University Bloomington law professor Jody Madeira told the Washington Post.
But many in law enforcement disagree. In March 2013, PoliceOne surveyed its 380,000 active-duty and 70,000 retired law enforcement officer members. Eighty-six percent of members believed that casualties from mass public school shootings could be reduced or “avoided altogether” if citizens had carried permitted concealed handguns in those places. Seventy-seven percent supported “arming teachers and/or school administrators who volunteer to carry at their school.” No other policy to protect children and school staff had such widespread support.
“A deputy in uniform has an extremely difficult job in stopping these attacks,” Sarasota County, Florida, Sheriff Kurt Hoffman told the CPRC. “These terrorists have huge strategic advantages in determining the time and place of attacks. They can wait for a deputy to leave the area, or pick an undefended location. Even when police or deputies are in the right place at the right time, those in uniform who can be readily identified as guards may as well be holding up neon signs saying, ‘Shoot me first.’ My deputies know that we cannot be everywhere.”
Similarly, Massad Ayoob, a self-defense advocate who has taught police techniques to law enforcement since 1974, noted: “When a life-threatening crisis strikes and seconds count, the real first responders are the citizens present.”
Past errors in the FBI Reports never corrected
“So much of our public understanding of this issue is malformed by this single agency,” notes Theo Wold, former acting assistant attorney general in the U.S. Department of Justice. “When the Bureau gets it so systematically – and persistently – wrong, the cascading effect is incredibly deleterious. The FBI exerts considerable influence over state and local law enforcement and policymakers at all levels of government.”
These omissions and discrepancies are not surprising given the limits of data collection and the judgment calls involved in categorizing such incidents. Law enforcement agencies around the country do not provide comprehensive reports of active shooter incidents, so local news coverage is a crucial source of information. The FBI contracts out this work to the Advanced Law Enforcement Rapid Response Training Center at Texas State University and then reviews and refines its findings.
The CPRC discovered cases the Center missed, but even the CPRC’s approach almost certainly misses incidents. “[T]here’s no reason to think that the [CPRC’s] list is complete, since there may well have been such incidents that weren’t covered in the news in a way that would come up on the Center’s searches,” UCLA Law Professor Eugene Volokh wrote in June regarding an earlier list of our cases that looked at just what otherwise would have been mass public shootings, a much narrower list than active shooting cases.
Asked about these discrepancies in August, the FBI declined to address them. A representative from the Advanced Law Enforcement Rapid Response Training Center, M. Hunter Martindale, suggested that its numbers were not definitive:
“We do appreciate you sending potential active shooter cases for the FBI team to review for inclusion in the active shooter dataset. As promised, I sent the email chain to the FBI team yesterday. As I’m sure you know, the FBI Active Shooter reports are released on an annual basis. My assumption is that any amendment retroactively adding cases would likely be included in a release with the annual report.“
Although collecting such data is fraught with challenges, some see a pattern of distortion in the FBI numbers because the errors almost exclusively go one way, minimizing the life-saving actions of armed citizens. “Whether deliberately through bias or just incompetence, the FBI database of active shooters cannot be trusted,” said Gary Mauser, an emeritus professor at Simon Fraser University in Canada who has extensively studied gun control and defensive gun uses. Mauser’s concern dovetails with those voiced by Rep. Jim Jordan in a July 27 letter to FBI Director Christopher Wray. Jordan alleged that whistleblowers have come forward claiming political biases in the FBI’s domestic terrorism data.
What is particularly troubling is the unwillingness of the FBI and the media to correct these omissions when informed about them. When Dr. John Lott worked at the US Department of Justice’s Office of Legal Policy and the Office of Justice Programs in 2020, the FBI was notified of their omissions involving potential mass public shootings, but they refused to correct those errors. Lott had previously alerted the FBI to similar problems back in 2015 and he published the list in the Academy of Criminal Justice Sciences Today in March 2015, but corrections were never made even after the FBI admitted they were missing cases.
When the CPRC emailed Ed White, the AP reporter who wrote that article, about the omissions in the Texas State numbers, he responded: “Our reporting, citing the specific research by Texas State U. over a 20-year period, was accurate. No correction was necessary.” The reporter did not need to take our word for these errors. A list of these cases and links to the news stories where mass public shootings were stopped was provided to him so that he could check out the omissions himself.
Requests to the Advanced Law Enforcement Rapid Response Training Center on exactly how much of their $66.9 million in grants from the US Department of Justice (here and here) were spent on putting together the FBI’s list of active shooting cases has never been answered.
This last video illustrates how the television media also uses the FBI to claim that attacks are not stopped by good guys with guns.