There have been 15 mass shootings in Arizona since 2013.
This is according to Mass Shooting Tracker, a crowd-sourced database that media heavy-hitters like Vox and the Washington Post have cited since a heavily armed couple opened fire Wednesday at a service center for the disabled in San Bernardino, California, killing 14 people and wounding 21.
According to official government statistics, however, there have been three mass shootings in AZ.
The difference between the two measures is even more dramatic on a national level. In 2015 alone, Mass Shooting Tracker reported 355 mass shootings. Government data shows 32.
So which is correct? It depends how a mass shooting is defined.
Historically, academics and — until recently — a majority of the media have counted incidents in which four or more people, excluding the shooter, are killed using a firearm. On the other hand, Mass Shooting Tracker, a website started by Reddit users after a mentally ill man unloaded a Bushmaster XM15-E2S rifle on a Connecticut elementary school in 2012, counts any incident in which four or more people are shot, regardless of how badly they are injured.
Some experts are alarmed by this shift in the way mass shootings are described in the popular press. But others see it as a good thing: a kick in the pants to motivate government officials to reform firearm laws.
The biggest problem with the emerging statistical approach, said Northeastern University criminology professor James Alan Fox, whose research has been influential in establishing the official definition of mass murder, is that media almost always roll out Mass Shooting Tracker's data in the wake of tragedies.
This leads the "unsophisticated reader," he said, to presume that every day, somewhere in the United States, there's a gunman with a chip on his shoulder brandishing an assault rifle, spraying bullets into a panicked crowd of strangers and killing multiple people.
In fact, Fox told New Times, that in a third of the incidents Mass Shooting Tracker has logged, no one died. In 95 percent of cases, there was one fatality.
Most mass killings involving firearms are not random or public, he said. More than half of cases are family-related. The next most common scenario with multiple victims is armed robbery. San Bernardino-style massacres are rare (last week's attack in California now is under investigation as a terrorist act with possible links to the Islamic State).
Of the seven incidents Mass Shooting Tracker recorded in Arizona in 2015, three stemmed from domestic violence, one was a dispute between two rival drug gangs, one was robbery-related, and one was a drunk altercation outside a Phoenix nightclub. Only one case, a shooting at Northern Arizona University in Flagstaff in October, might be defined as a public shooting. But even though one person was killed and three more were shot, as many argued in its aftermath, it didn't fit the typical profile of a crazed gunman executing a carefully planned attack on r andom students (it appeared to be a heat-of-the-moment altercation outside a fraternity house).
"I wouldn't and I don't ignore family or felony-related murder. When six people are killed in their home, that is no less tragic than when six people are killed in a workplace or a restaurant," Fox said. "But public shootings scare people most because the feeling is they can happen at any time or any place — no matter who you are or what you are doing."
Failing to differentiate between types of shootings "scares people out of proportion with the risk," Fox said.
"When people are afraid, they don't necessarily respond in the right way," he said. "If some policy or procedure is a good idea today in the wake of a shooting,
like the one we had in San Bernardino, it will be a good idea in two months when emotions aren't so raw."
Because Mass Shooting Tracker and others that have attempted to log the number of shootings, such as Mother Jones and former New York City Mayor Michael Bloomberg's activist coalition Everytown for Gun Safety, only have started counting within the last few years, Fox said, their numbers lack historical context. To fill in information from previous decades, compilers have searched archived news reports, which get less comprehensive as you go back in time. As a result, he argued, their statistics inaccurately suggest massacres are on the rise.
By contrast, the government has kept statistics on firearm-related homicides with four or more victims for decades. Because the data was collected as the shootings happened, he argued, it's much more reliable. According to Fox's analysis of this data, the number of incidents involving firearms has averaged about 20 a year in the United States for decades with no statistically significant increase, he said.
"The only thing that's really increased here is fear," he said. "A lot of that has to do with a total misunderstanding of these numbers."
Sherry Towers, a research professor at the Simon A. Levin Mathematical, Computation, and Modeling Sciences Center at Arizona State University, acknowledged the holes in data collected by organizations like Mass Shooting Tracker, but she argued that it's still important to analyze gun violence separately from mass murder. Though they may be flawed, she said, informally collected numbers are the best researchers have to go on because Congress passed a law in 1996 banning the Centers for Disease Control from funding studies that may be used to advocate for or promote gun control.
Although most academics still rely on the government's mass murder counts, she recently used a data set similar to Mass Shooting Tracker to conduct an analysis of how media coverage affects gun violence. She found that, for 13 days following a highly publicized incident with four or more victims, the probability that another such shooting will occur is heightened. About 20 percent to 30 percent of mass shootings happen during these periods of contagion.
The discovery suggests, she said, a different narrative from the one laid out by looking at death tallies alone. It appears, she said, that people see news about an event and "get an idea to go on a shooting rampage."
In addition to being helpful to researchers trying to study gun violence, statistics serve an important role in educating the public and inspiring policy reform, Towers argued.
"I think that most Americans, unfortunately, are unaware of how often mass shootings happen," she said. "They literally are happening on average at least once a day [in the U.S.], but the stories almost never make it past the local news."