Tracking the sound of gunshots: Is ShotSpotter the solution to Toronto gun violence?

An audio surveillance system that detects gunfire — tentatively greenlit by Toronto city council — poses privacy concerns as well as the risk of overpolicing already stigmatized neighbourhoods, a professor of sound studies says.

Toronto city council approved funding for ShotSpotter and other anti-gun violence measures on the heels of a spate of fatal shootings in Toronto, including a mass shooting in Toronto's Danforth neighbourhood. The California-based technology uses microphones to detect the sound of gunshots and triggers an alert to notify police.

"Torontonians are in shock, without question, but that's not good municipal policy," said Lilian Radovac, an assistant professor at the University of Toronto's Institute of Communication, Culture, Information and Technology.

"It can be one thing to try to address the tragedy and the grief and the trauma that Torontonians are feeling...that's a big part of the role of a city councillor. But that's a very separate question from whether or not a technology ... is actually going to work to address the core concerns," she told Day 6 host Brent Bambury.

So far this year, there have been 228 shootings in Toronto and 29 of have been fatal. 

ShotSpotter is used by dozens of cities in the U.S., but its track record is still unclear.

On the morning after the mass shooting in Toronto, Radovac wrote to her local city councillor expressing her concerns about the tool.

She spoke with Bambury about some of her issues with the technology. 

Brent Bambury: Immediately after the shooting you wrote to your city councillor, about ShotSpotter. Why?

Lilian Radovac: I am very nervous about this system and I was concerned that Toronto was on the precipice of adopting a software that I knew a little bit about and had some reservations about. And then that just went into high alert mode after the shooting because I knew that councillors would be feeling a lot of pressure to do something.

BB: You're not alone in having reservations, but Toronto is not alone in thinking about this system. Two of the largest cities in the United States — Chicago and New York — have adopted this technology. In Chicago, it was expanded last year. Are those two jurisdictions — along with the 85 others that have purchased the system — are they making a mistake?

LR: I know for a fact because I lived in New York for two years, I lived in Brooklyn, that there are a lot of people in that city that have concerns about over-surveillance of certain communities — especially in Brooklyn and the Bronx. I know Chicago probably shares similar concerns, but I also know that both have city mayors who have pushed very hard for these kinds of surveillance technologies. But there [are] definitely critiques on the ground to be sure.

BB: Well, let's look at your critiques because sound studies is your field. What have you learned about the effectiveness of the ShotSpotter technology?

LR: Well, the first thing I learned was that there is very little public information about it. ShotSpotter data, which it collects through placing microphones in targeted communities, is kept private and that's proprietary. So it's not released to the public for people like me or journalists to analyze.

But what I did come across was some great reporting by Matt Drange, who at that time was an investigative journalist with Forbes magazine. So, you know, a decent reputable magazine — not a small rag — who managed to get information from what he calls customer cities. And he evaluated that information and discovered that very few arrests resulted from the alerts that were produced by this technology. There were a whole lot of alerts, but very few cases of actual summonses or arrests that came out of this.

BB: But the claim that the people who develop this technology make is that this can pinpoint the location of a shot fired and do so rather quickly. So arrest is just one metric. An alert might be another important metric. If police can get to the position of a shot faster because of this technology, is that not valuable?

LR: And that is, in fact, the main argument: that it reduces the time that it takes for a police officer to get to a potential crime scene. What's interesting is that — and this is according to Drange's analysis — depending on the jurisdiction, between 30 and 70 per cent of the alerts that went in and were in fact referred to police, brought the police officers out, led them to investigation and they found no evidence of any shooting or, in fact, in some cases of any crime at all.

The problem is that when you bring a police officer into a place for an investigation — remember, they're being told, 'This is a probable gunshot' — their adrenaline is probably up. They go into a community looking for criminals. And that means that they're going into over-policed communities. And if the rate of error — a rate of alerts that turn out to be nothing — is as high as 30 to 70 percent, you're looking at a lot of unnecessary police investigations in neighbourhoods that are already over-policed and targeted and profiled.

BB: I want to go back to the missing evidence, the fact that you said earlier that there is not enough data for you to be able to make a decision based on the science. Why do you think that evidence is being suppressed?

LR: It's a matter of monetizing that information. So it's not suppression in the sense that we would think like, 'Ooh, scary political suppression' or something like that. Not at all. But these are private tech companies who are monetizing this knowledge to sell a product. So they want to be able to control that information; they want to be able to control the analysis.

I think that's probably the biggest concern right now, and the thing we should be lobbying for, to make sure that public researchers, including journalists, are able to access this information and truly evaluate whether or not it's effective.

BB: There are also the concerns of privacy abrogations in the regions where the technology is being installed. Do you think that that is a reasonable concern?

LR: I think privacy is a huge issue. In fact, I think it's the one we tend to talk about most frequently when we talk about city tech. And I think that's a very interesting question when you look at targeted communities. Because then you think about unequal notions of privacy. Does somebody in the Annex have a greater expectation of privacy than someone living at Jane and Finch? And that's a very interesting entry point.

BB: This is a city that went through a lot of pain this week. Three people are dead. Many people are wounded. There was a shooting in a residential neighbourhood at a time when people should be out with their families, enjoying themselves. That shooting happened on Sunday night. City council voted on Tuesday. What was lacking in the conversation, on ShotSpotter, in that duration of time?

LR: It can be one thing to try to address the tragedy and the grief and the trauma that Torontonians are feeling. That's valid, that's important and that's a big part of the role of a city councillor. But that's a very separate question from whether or not a technology that will cost us $4 million is actually going to work to address the core concerns. The evidence that we have to this point, which is paltry, admittedly, says it won't help. Of course we're going to have emotional responses. These are horrible tragedies. Torontonians are in shock, without question, but that's not good municipal policy.

This interview has been edited for length and clarity. To hear the full conversation with Lilian Radovac, download our podcast or click the 'Listen' button at the top of this page.