In the wake of the Wednesday shooting at a Charleston, S.C., church that left nine people dead, John R. Lott Jr., the prominent researcher whose work is embraced in gun-rights circles, offered a theory of the gunman’s motive.
“Churches, like the one in Charleston, preach peace,” he wrote, “but the killer there probably chose that target because he knew the victims were defenseless.”
It’s a familiar pattern: In the immediate aftermath of the 2012 massacre at an Aurora, Colo., movie theater, conservative commentators were just as quick to reach a decisive verdict on the killer’s intentions, blaming the bloodshed squarely on the Cinemark theater’s no-guns policy, which they argued left moviegoers defenseless. “Was there nobody that was carrying that could’ve stopped this guy more quickly?” Texas Republican Rep. Louie Gohmert asked, before pointing to an incident of domestic dispute in Tyler, Texas, in which a gun-carrying citizen intervened. Lott postulated further that “gun-free zones are a magnet for those who want to kill many people quickly.”
The perceived vulnerability of so-called gun-free zones is more than just a gun-rights talking point: It has emboldened a wave of legislative action on both state and federal levels to expand the freedom to carry firearms in previously forbidden places, such as schools or private businesses. Last year, Georgia enacted a “guns everywhere” law that has dramatically expanded where firearms are permitted in the state, including bars, classrooms, and government buildings. Just this month, Texas passed a bill extending the concealed carry of firearms onto college campuses, a move that Texas Lt. Gov. Dan Patrick applauded, saying, “I am proud of the fact that the Texas Senate is making history while defending life, liberty, and our Second Amendment right.”
At first glance, some mass shootings do appear to validate a central claim of gun advocates — namely, that killers strategically target gun-free zones because they expect weak resistance from unarmed civilians. But an analysis of the best available research points in the opposite direction: There is no evidence that mass killers select locations based on gun policy, or that lawful gun owners have been able to intervene to stop these attacks.
In the Aurora case, the shooter gave no indication that the theater’s gun-free policy played a part in his motives. His personal journal, made public during his ongoing trial, contains not even a cursory mention of gun-free zones or the consideration of armed civilians, but instead details a more pressing concern about how to attack the “isolated, proximal, large” space: finding the right parking spot.
John R. Lott is, by a significant margin, the most prolific and outspoken researcher on gun-free zones in the United States. He is the author of More Guns, Less Crime and a vociferous opponent of gun laws, having frequently testified before Congress in favor of expanding Right to Carry (RTC) policy. Lott abhors the idea of gun-free zones, and has made the ambitious claim, on numerous occasions, that “with just two exceptions, every public mass shooting in the United States since at least 1950 has taken place where citizens were banned from carrying guns.”
Central to Lott’s argument against gun-free zones is a 2000 study in which he claimed to find that the expansion of RTC laws reduces the number of people in those states killed or injured in multiple-victim shootings by a staggering 78 percent. Lott’s study, however, suffers from enormous flaws, including incorrect statistical modeling and dubious data-selection methodology.
In one example of statistical malpractice, Lott excludes many mass-shooting incidents in which the shooter was committing an additional felony (such as armed robbery) during the crime, despite the fact that felony-related mass murders account for 36 percent of the data set on which he bases the study. Lott’s explanation for doing so was an unjustified presumption that bystanders in crimes like robberies or drug deals will already “be engaged in unlawful activities that often require them to carry guns.” However, analysis of this claim reveals that 69 percent of the mass shootings excluded by Lott involved robberies committed in public locations (like convenience stores and fast-food restaurants) where the bystanders were innocent civilians. If RTC laws are to have any effect at all, then surely they would apply to such situations, making it unclear how Lott could choose to ignore them.
When Lott’s research is compared to a more recent study using more appropriate statistical models and a wider range of available data, the beneficial effect of Right to Carry policy vanishes. The authors of a 2002 study, a trio with combined criminology and economics expertise, evaluated RTC laws in 25 states from 1977 to 1999, an expanded version of Lott’s analysis (which covered 23 states in that same time period). They concluded that “RTC laws have no effect on mass public shootings at all.”
Perhaps the most glaring flaw in the argument against gun-free zones, in the context of mass shootings, is its underlying assumption that shooters are rational actors. Lott himself admits that about half of criminals who commit mass shootings have received a “formal diagnosis of mental illness,” yet his model requires them to act precisely as we know they don’t: as hyperrational, calculating machines, intentionally seeking out gun-free environments for the sole purpose of maximizing causalities.
In reality, many shooters target a location based on an emotional grievance or an attachment to a particular person or place. An FBI study of 160 active shootings (defined as a shooter actively attempting to kill people in a populated area, regardless of the amount of fatalities) between 2000 and 2013 — including the high-profile mass shootings in Tucson and Aurora — shows that of the shootings that occurred in commercial or educational areas, the shooter had some relationship with the area in 63 percent of the cases.
This was the case on October 21, 2012, when a gunman barged into the Azana Day Salon in Brookfield, Wis., searching for his estranged wife, who had recently filed a restraining order against him. Despite the order, he still managed to purchase a firearm through an online source. The man murdered three people and wounded four others before finally turning the gun on himself.
While active shootings have similar characteristics to public mass shootings, they are defined by the activity of the shooter rather than the casualty count. Hence, active shootings include incidents that were stopped before the “four or more deaths” threshold to reach the commonly used classification as a mass killing with a firearm. This is extremely useful from an analysis standpoint, as it allows for a more complete picture of which public places shooters are targeting, their potential motivations, and what actions could be taken to reduce casualties.
Of the 39 shootings in the study that occurred in educational environments, 31 of the shooters had some relationship with the school (27 were current or former students). Out of 23 businesses with no pedestrian traffic (i.e., private offices rather than stores) where shootings occurred, 22 of the shooters were current or former employees. These shooters are overwhelmingly motivated by some grievance rather than a desire to maximize casualties, which makes it highly unlikely that a gun-free policy had any bearing on the choice of target.
The exceptions to this pattern are shootings that occur in open areas and businesses with pedestrian traffic, such as the shootings in Tucson and Aurora. Lott would argue that both of these location types are prime targets for shooters intent on maximizing casualties, as he did in a Chicago Tribune column in which he encouraged citizens to arm themselves before visiting the Mall of America to protect against the threat of terrorism, in defiance of the mall’s rule barring shoppers from carrying guns.
But counter to what gun-rights advocates claim, many of the active shootings in the FBI’s database occurred in areas that were not gun-free zones. Our own analysis of the FBI study, in which we looked at all 160 incidents, examining state and local laws along with the firearms policies of individual businesses, found that of the 65 shootings in open spaces and businesses with pedestrian traffic, at least 25 occurred in areas permitting firearms.
That may amount to less than 50 percent — but we counted only those cases in which the evidence indicating that the open space or business was not a gun-free zone is unmistakably clear. Complicating this analysis are differences in state laws governing whether “no-firearms allowed” policies in businesses are actually enforceable, and the fact that many concealed carriers aren’t aware of (or deliberately ignore) no-firearms policies.
For example, out of six mall shootings included in the report, two occurred in malls with gun-free signs, and yet armed citizens attempted to intervene. In the Clackamas Town Center shooting in Oregon, a permit holder confronted the shooter (but did not fire) at the end of the rampage; in the Trolley Square Mall shooting in Salt Lake City, an off-duty police officer helped subdue the shooter. Both men were in explicit violation of gun-free policies — but their presence means that for the shooters’ purposes, those malls were not gun-free zones after all.
A study conducted by Everytown for Gun Safety further corroborates the active-shootings data by examining every mass killing with a firearm between January 2009 and July 2014. (The Trace received seed funding from Everytown for Gun Safety.) Of the 33 incidents that occurred in public places, 16 took place in part or wholly within areas where guns could be legally carried. Two more incidents occurred where an armed guard or police were in the immediate vicinity.
Gun-rights advocates have long defended the public carrying of guns on the basis of a widely debunked 1992 study estimating 1.5 million to 2.5 million incidents of defensive gun use per year. Armed citizens, they argue, can uniquely limit the damage of a would-be mass shooting by stopping it before it escalates. Empirical evidence, however, indicates that defensive gun use may occur far less frequently: According to the Gun Violence Archive, there were only 1,600 verified accounts of defensive gun use in 2014. The best current statistical model, which corrects for numerous weaknesses in Lott’s body of work — including coding errors (in his 2000 study on Right to Carry laws, both Philadelphia and Idaho had their “year of adoption” dates for concealed-carry laws coded incorrectly), a lack of standard criminal-justice controls, and a lack of clustered standard errors (a standard econometric practice) — suggests that concealed-carry laws may actually increase the rate of aggravated assaults.
The same FBI report analyzing active-shooting incidents from 2000 to 2013 decisively refutes the idea that average civilians would be likely to stop would-be mass shootings if only gun-free zones didn’t stand in the way. By focusing on active shootings rather than mass shootings, we are able to test the hypothesis that the reason so few mass shootings are prevented by permit holders is because the incident is stopped before the casualty count reaches the mass-shooting threshold of four victims. The report found that of the 160 incidents, only one active shooting was stopped by a concealed-carry license holder: a U.S. Marine. In comparison, four shootings were stopped by armed guards and two by off-duty police officers. Twenty-one were stopped by unarmed civilians.
In response to this report, Lott issued an error-filled analysis accusing the FBI of playing politics and manufacturing an upward trend in mass shootings. However, Lott’s critique ignored the entire point of the report, which was to study active-shooting incidents (not mass shootings). As the authors explained in a rebuttal earlier this month: “Lott’s essential argument is a straw man; he accuses us of saying something that we did not and then attempts to show this is wrong.”
Lott and other gun advocates also frequently reference incidents that were rightfully excluded by the FBI authors because they bear no resemblance to an actual active-shooting situation. To buttress his case, Lott maintains a highly selective list of incidents that, as he explains, “only includes cases where mass public shootings were stopped.” But this roster of prevented “mass public shootings” includes two knife attacks (neither of which had any fatalities), a Marine firing his gun in an empty parking lot, several robberies (a type of incident Lott excludes from his statistical analysis), and, under “possible cases,” the Tacoma Mall shooting, in which a permit holder confronted the shooter and was swiftly gunned down and subsequently paralyzed for life.
When a “good guy with a gun” does intervene in an active shooting, things can go terribly awry. On June 8, 2014, an armed couple burst into a CiCi’s Pizza in Las Vegas screaming, “This is the start of a revolution!” They quickly gunned down two police officers eating lunch, and then moved to a nearby Wal-Mart. One customer, a concealed-carry license holder, drew his gun rather than flee, but was immediately shot. As it would turn out, all three of the couple’s victims that day were armed.
Another example: On Jan. 8, 2011, a gunman opened fire on an outdoor meeting between Rep. Gabrielle Giffords and her constituents in Tucson, Ariz., killing six and wounding 13. When the killer was forced to reload, he was tackled by a bystander. Having heard the gunshots, an armed man ran to the scene. He saw two men wrestling and assumed the wrong man was the shooter. Had it not been for other bystanders quickly correcting him, he could have ended up shooting the wrong person. Afterwards he stated: “I was very lucky.”
[Photo: Flickr user John Loo]