| |

Is Statistics Canada Overstating ‘Firearm-Related Homicides’?

Statistics Canada Firearm Homicide Flawed Data Guns

(Update 14 May 2018: Adds data on automatic firearms and “firearm-like weapons” in ninth paragraph.)

TheGunBlog.ca — Statistics Canada’s annual homicide report counted 223 “firearm-related homicides” last year. Data tables accompanying the publication show 30 of them, or 13.5 percent, may not involve guns at all.


  • StatsCan counts flare guns, nail guns and pellet guns as firearms.
  • StatsCan may be overstating gun-related deaths, or may not. We don’t know.
  • Opacity and uncertainty mean we can’t trust stats on gun-related crime, and government may be using bad data in policymaking.

Firearms + Non-Firearms = ‘Firearms’?

StatsCan’s possible overstatement of gun-related killings stems from the agency including a mishmash of firearms and non-firearms in a single category of “Other firearm types,” and counting all of them as “firearms.”

The details are in footnotes to Table 3 and Table 5 of the “Homicide in Canada, 2016” report, published yesterday. The notes say the “Other” category includes “firearm-like weapons,” such as flare guns, nail guns and pellet guns, which typically aren’t considered firearms, as well as “fully-automatic firearms,” which are. It also includes homemade “zip guns” and “firearm-type unknown.”

Beyond eroding trust in gun statistics, the potentially flawed data could be used by the government to force more restrictions onto law-abiding gun owners, even as it prepares to unveil a set of new laws. Activists hostile to gun owners and close to the Liberal Party-led government are already citing the questionable statistics to push for more controls.

The homicide report is based on police reporting, and the “Other firearm type” category allows police to count anything as a “firearm” if it was used as a weapon and launched a projectile.

Fires Projectile = ‘Firearm’?

“The purpose behind having an ‘other’ category is to offer an opportunity to provide a category that allows police services the option to identify incidents where the homicide is committed using a weapon that fires a projectile that does not fit the traditional definition of a firearm,” Statistics Canada said by e-mail yesterday in response to questions from TheGunBlog.ca.

“The reference to nail guns, flare guns and pellet guns is made to provide examples of firearm-like weapons,” the statistics office said.

StatsCan said it doesn’t show details of the “Other” category. The opacity results in uncertainty about whether it includes guns or not.

A separate table shows that “firearm-like weapons” were used in zero homicides in 2016. The table also shows that automatic firearms were used in 6 homicides, and that police couldn’t identify the type of firearm or “firearm-like weapon” in 24 cases and marked them as “unknown.”

Exaggeration, Imbalance

StatsCan exaggerates levels of “firearm-related” crime more generally by including incidents where a gun was found nearby, even if it wasn’t used to commit the crime, according to research by Gary Mauser.

The homicide report counted at least 193 killings using bullets last year and 175 deadly stabbings. It mentioned the word “firearm” or “firearms” 30 times and the word “knife” or “knives” zero times. Knives were used in 90 percent of stabbing homicides, StatsCan told TheGunBlog.ca.

The report has expanded its analysis of shootings to help policymakers assess risks and develop policies, StatsCan said a year ago after publishing the 2015 edition. Firearms were mentioned 32 times in the 2015 edition, 19 times in 2014 and 28 times in 2013. None of the reports mentioned knives. Stabbing homicides exceeded shooting homicides in all three years.

‘Firearm’ Injuries Déjà Vu

The Liberal Party, which developed the current regime of laws hostile to legal gun owners when it was in government in the early 1990s, campaigned in 2015 on promises to punish lawful shooters and retailers with additional restrictions.

Earlier this year, research claiming to study firearm injuries in children and youth, and published in the Canadian Medical Association Journal, was undermined by a series of flaws, including counting non-firearms as firearms.

Decades of Bad Data?

StatsCan’s series of possibly bad data goes back to at least 1996, and the 30 “Other firearm types” counted in 2016 homicides were the highest since then, in both number and percent of the total. Any inaccuracies would also lead to errors in StatsCan’s calculations of the rate of gun-related killings per 100,000 people.

Even if none of last year’s “Other” turn out to be firearms, it would still mean 193 people were killed by bullets fired from guns, more than in 2015, according to StatsCan. A surge in gang shootings in Toronto and Vancouver was behind the increase.

© 2017 TheGunBlog.ca