Context Matters: The S.W.A.T. Study
A while back I was sent a D.M. from a policing friend on Twitter. He wanted to know my opinion on a recently released study purporting to show that Canadian policing was becoming increasing militarized. Somewhat unusually for any Canadian criminology research, it got picked up by the media (in large part because the authors vigorously promoted it through news releases, interviews and opinion pieces). Curious about the central argument, I did a deep dive into the study and here’s what I found.
The authors identified five sites – the Vancouver, Calgary, Winnipeg and Ottawa Police Services and the Ontario Provincial Police – and, utilizing Freedom of Information requests (FOI), asked for numerical data on Tactical Unit deployments from 2006 to 2017 (although not each year, just selected years). From this data, they drew the following conclusions:
Canadian policing (as exemplified by 4 municipal agencies and one provincial service) is becoming more akin to American style policing with an increased emphasis on military style equipment, training and deployments.
Canadian police agencies are now more routinely deploying Tactical Units to attend to ‘normal’ or ‘routine’ policing calls.
Here are just some of the issues I found with the study:
It included 4 of the larger municipal services in Canada out of a possible 200. It did not include medium to small agencies, which, if anyone understand policing resources in Canada, would be expected to change the results.
Large chunks of Canada are policed by the R.C.M.P. As we know from the Moncton Review, this means that large chunks of Canada have been policed by a significantly under-equipped police service.
Reliance on F.O.I. data without interviews or other data to provide context means the paper evidences a serious lack of understanding of the operational environment in which Tactical Unit (TU) members work. To illustrate:
Tactical Unit members in many agencies across Canada are regular police officers deployed to perform regular patrol functions. They aren’t typically sitting around pumping weights, stroking carbines and waiting to be called to go to a speed trap on the off-chance that Al Capone is driving around. This means that it is very likely that some of the call data might reflect routine practices because TU members were involved in routine policing activities that went a bit sideways.
The authors’ definition of ‘routine’ policing activities includes some actions that an experienced police officer might recognize as potentially fraught with peril, including serving warrants and certain vehicle stops. What the police agency knows, and the researchers don’t (because all they had were the numbers), is what the circumstances of the call were, including whether they were serving a high-risk individual with a history of violence, or whether the vehicle stop involved someone who was trying to ram his or her car at police. Having years worth of coroner’s records on police-involved shootings, many of which involved cars being used as weapons, I can tell you that the notion of all vehicle stops being ‘routine’ is a fiction.
I’ve now had a chance to look at some of the original data the researchers were provided (3 of the agencies provided it to myself and Craig Bennell (Carleton) for a follow-up analysis; 1 agency refused to participate in any way and another has yet to respond to inquiries). What would a new study look like? It would include CONTEXT. As in the present case, numbers alone often cannot tell you the complete story. The complete story entails interviewing TU members about their experiences in order to make those numbers make sense in a way that is more accurate and thus more reflective of what is actually going on (rather than solely reliant on the researchers’ interpretation, which may be slightly biased).
And therein lies the crux of what it means to be evidence based in the context of evidence-based policing: the best research combines the researcher’s knowledge with police experience to produce well-informed, actionable research.