top of page

Verify, verify ... and then verify

I have a strange collection on my hard drive. It consists of coroner’s inquiry records for all of the police-involved homicides from the provinces of Ontario and B.C. going back to about 2000 and up to 2015. The thing I like about coroner’s records is they provide not only some useful context for understanding a death event, but they also give you a reasoned judgment on how that event should be labeled: natural, accidental, suicidal, homicide or unknown. Deaths that are deemed to have occurred through the actions of another person are a homicide. Unlike American TV shows, where homicide is another name for murder and the excuse for Armani-clad, beautiful people to run around in Hummers, Canadian coroners’ findings that an occurrence is a homicide attach no criminal blame. It is just a fact.

Having collected and extensively read through these files, it was therefore something of a surprise when CBC News (2018) released research they claimed detailed 461 “Fatal encounters with police in Canada: 2000-2017”. In case one had any doubt as to what these ‘fatal encounters’ entailed, subsequent articles left little doubt that CBC was defining these events as “Canadians who died at the hands of police” (Marcoux 2018), or what the Coroners’ records might term ‘homicide.’ What was unsettling to me was that I had all of the Coroner’s records for the two provinces in which the majority of police-involved homicides occur, plus SIU and other reports for the provinces of Alberta and Manitoba, and I had nowhere near 461 cases.

Intrigued, I paid a visit to the online archive of CBC cases (this can be found at: and started scrolling through the photos. One thing that immediately struck me was the face of a young, white woman – an individual from a demographic group that tends to be under-represented within police-involved deaths. The file was on a woman called Nadia Racine. Here’s the CBC summary of her ‘fatal encounter’:

Nadia Racine’s neighbour called 911 after he heard screaming and banging coming from her apartment. Police forced their way into Racine’s apartment when no one would answer the door. According to evidence presented to Quebec’s independent investigation bureau, officers found Racine agitated and banging her head. They placed her in handcuffs before realizing her condition appeared to be “deteriorating.” She was taken to hospital, where she died. The cause of death remains under investigation.

How exactly self-harming behaviour engaged in independently of police actions was determined to be a ‘fatal encounter with police’ is unclear. However, leaving that aside for the moment, what particularly troubled me was that the ‘cause of death remains under investigation’ – meaning, that there had been no actual finding of homicide and that this case might very well be labeled a suicide and therefore not exactly evidence of someone having ‘died at the hands of the police.’ In fact, if the whole story is true, then the officers involved actually tried to save Nadia’s life.

Nadia Racine’s story is not unusual in the fact that, despite the case was still being under investigation, it was included as an example of a ‘fatal encounter’, ‘deadly force’ and/or a death ‘at the hands of police.’ The inclusion of cases in which no determination had yet been made as to how the person died is sloppy at best and deceptive at worst.

Scrolling through the 461 cases cited by CBC also reveals some other interesting facts about their research strategy. Included among them are 23 ‘unknowns’ – that is, reports of a police-involved death with little to no confirmed details, such as the individual’s identity. Given that several of the cases occurred in Ontario, these could easily have been matched to Coroner’s records or SIU reports, which would not only have confirmed any details, but provided a name and a finding as to how that individual died. That CBC’s researchers included these cases without apparently cross-checking them against other sources, or included them in cases when there was no supporting corroborating information, is problematic. If one of my graduate students had done that, those cases that could not be verified against other sources would be removed. Here’s an example of one of these ‘unknown’ cases:

A police officer had no reason to suspect a man she arrested for public drunkeness [sic] at a Tim Hortons was concealing baggies of marijuana in his mouth when she left the man in her cruiser, an independent investigation found. When the officer returned to the squad car she found him banging his head against the door before becoming unresponsive. Doctors found his airway completely blocked with the baggies, and he died the following day from complications due to suffocation. No charges were laid.

Again, contra the image of a gun muzzle being pointed directly at the reader, which was used to illustrate the CBC story, this is yet another case where someone appears to have died not at the ‘hands of the police’ but by their own hands.

Undoubtedly, were anyone else to go through each of the 461 cases they would similarly find a number of them not meeting anyone’s criteria for homicide or, as the CBC likes to put it, ‘deadly force’ by police. But here’s the thing: when the CBC published the series of articles from their research, there were very few voices challenging the methodology they employed and their inclusion criteria. I should know, I was one of them. Rick Robson from the London Police Association was another, as was Tom Stamatakis from the Canadian Police Association. None of us needed years toiling in an Ivory Tower on statistical formulas or some special egghead hat to figure out there was something flawed about the ‘research’ presented. We just needed to apply some logic and a willingness to verify (or, more accurately, falsify) what we were reading.

At the end of the day, that’s all that research methodology is: a logical, reasoned approach to understanding a phenomenon.

bottom of page