Unhelpful Exaggeration of Data on “School Shootings” Sets Back Scholarship on Gun Violence

Coincident with a murder-suicide at a high school outside of Portland, Oregon on June 10th, some “data” on school shootings began to circulate on the internet. The “data” came from Everytown for Gun Safety, a gun control advocacy organization that brings together Mayors Against Illegal Guns and Moms Demand Action for Gun Sense in America, all of which are heavily funded by Michael Bloomberg (more on the significance of that later).

The raw “data,” presented in a table on the Everytown web site, were transformed into a graphic by Zee Maps and promoted on the Twitter feed of Huffington Post deputy managing editor Mark Gongloff.  The map was later published on the Huffington Post website under the title, “There Have Been 74 School Shootings Since Newtown.”

Courtesy of Huffington Post

The graphic was picked up by untold numbers of news outlets, including the UK edition of Huffington Post’s Twitter feed:

Tweeted on June 11, 2014

Soon enough, critics of gun control began interrogating the “data” underlying this map, led by freelance writer Charles C. Johnson on his Twitter feed. Johnson pointed out that many of the “school shootings” included in the Everytown list, and reproduced in Gongloff’s map, do not jibe with common understandings of what a school shooting is (e.g., suicides, shootings on school grounds in the middle of the night). And then came a back and forth of criticisms and defenses of the “data” and map.

The Tampa Bay Times’ Pulitzer Prize winning website, Politifact.com, has since stepped in and applied its “Truth-o-Meter” to the claim of 74 school shootings since Sandy Hook. Its judgement: MOSTLY FALSE.

Before Politifact beat me to the punch, I spent some time asking people what comes to mind when I say “school shootings.” Not surprisingly, they all conjured up images of an active killer shooting random victims in a school, sometimes colleges or universities, but more often elementary schools or high schools. Basically, scenarios like Sandy Hook, Columbine, Virginia Tech, or the recent shooting in Oregon.

Politifact concluded that only 10 of the 74 shootings fit with these common understandings. Of course, this is 10 too many. Indeed, no exaggeration is needed to say that this is a problem. Yet as a political advocacy organization, Everytown has a vested interest in exaggerating the threat of gun violence in schools, and news media organizations like Huffington Post have a vested interest in running scary headlines (see my previous comments on Michael Glassner’s culture of fear argument).

That political advocacy organizations and the media cultivate a culture of fear is business as usual. But it has some broader pernicious effects. The conflation of homicides and suicides, killing of innocents and drug related murders, and so on, plays into the sense (justified in many cases) that gun control advocates do not understand — or even seek to understand — the roles of guns in American society. They certainly do not understand the positive role, and they don’t even understand negative role. Indeed, they often don’t think guns should have any role and this becomes the driving force for their work.

In addition, I look at these situations largely from the perspective of a scholar, and what I see is not good. Blatant use of bad data like this delegitimizes good data. Indeed, it delegitimizes all data. It contaminates the work of scholars who have nothing to do with the misuse of data by political advocates. It plays into the sense (justified in many cases) that scholars who study gun violence are working from a pre-determined position that guns are bad and the only question is how bad and what specific measures are needed to control them.

I would hope that Bloomberg-funded public health researchers at Johns Hopkins would jump in to advise Bloomberg-funded advocates at Everytown for Gun Safety on the honest and appropriate use of data on gun violence. If these gun researchers would call out gun control advocates who behave badly with data, they might increase their credibility as objective scholars and make it easier for all of us to do our work (and maybe even get funding for it).