Detroit Police Chief Todd Bettison wrote an opinion piece in these pages on June 18. (“ShotSpotter isn’t magic, but in Detroit, it’s working,” Detroit Free Press, June 18.) Among other things, he stated the evidence is clear that ShotSpotter, an acoustic surveillance gunfire detection technology, is functioning effectively as a tool in the city’s public safety strategy.
Unfortunately, the data do not support the claims of the chief.
ShotSpotter is an error-prone, flawed and risky tool. Soundthinking, the company that operates ShotSpotter, claims a 97% accuracy rate with “extremely low false positives” across all their customers.
Many of their customers have studied the technology within their city limits and discovered a completely different story. The New York City Comptroller’s office did a well-resourced audit of the technology and found that just 8% to 20% of ShotSpotter alerts were confirmed to be actual shootings.
Chicago’s Office of the Inspector General concluded after significant quantitative analysis that ShotSpotter alerts rarely produce evidence of a gun-related crime or lead to other crime-related evidence.
Cities like Charlotte, Dayton, Nashville and Atlanta have similarly found the technology to be faulty and ineffective. The Associated Press also conducted an investigation of the technology, finding the system often “can miss live gunfire under its microphones, or misclassify the sounds of fireworks or cars backfiring as gunshots.” ShotSpotter employees can and do modify the data ShotSpotter collects, according to court records. That means someone outside of our city is helping make the decision whether or not to dispatch our officers ― and charging us a pretty penny to do so.
The chief’s rebuttal to critiques of ShotSpotter is that Detroit uses this technology differently. To date, the police department has not publicly indicated a unique way we are using the technology.
Many of ShotSpotter’s clients and former clients also choose limited deployment areas and pair the technology with other interventions. Like many of you, I believe Detroit is the greatest city in the world, but our utilization of ShotSpotter is no different than in Chicago, Atlanta or New York.
There are many reasons to be skeptical about the use of this technology. The price for this nine-month contract is over $2 million dollars.
I believe most Detroiters would have strong opinions of other ways we could spend such a large amount of our tax dollars. When in a similar position, that’s what residents and officials in the City of San Antonio did. While discussing their cancellation of the technology, their police department and council described their thinking as follows:
“It doesn’t make the community feel safer, it doesn’t reduce the number of gunshots in our community … It doesn’t prevent you from being shot. We’re going to use that money to provide more community engagement, which ShotSpotter [now Soundthinking] can’t provide and I believe it will … resonate with the folks who live in those areas a lot more than having some alert go off and police come after the fact.”
ShotSpotter also presents real privacy risks. The microphones used by the technology to purportedly triangulate gunshot positions are live and recording at all times. Their placement can be on your block, near your home or outside of your place of worship. At a recent city council meeting, Deputy Chief Mark Bliss said that the sensors cannot pick up voices because they are only activated when the sound hits the sensors.
However, a recent audit of the technology by NYU Law shows it to be entirely possible for Soundthinking to capture individual voices near its sensors and use them for targeted voice surveillance.
Soundthinking itself has provided recordings of individual voices from their sensors in court proceedings. There isn’t a question of whether the microphone can record a human voice.
The only question is if and how that improperly collected data can be used against residents of the city, in violation of their rights to privacy.
If the department is truly data driven as Chief Bettison says, it’s time to drop this technology.
One of the only studies you’ll find touting the technology’s effectiveness was paid for by the company.
It should be no surprise that they are making huge claims when trying to win a $2 million contract. As good stewards of our tax dollars, we must consider an admission from one of the company’s analysts, when asked about their accuracy guarantees, who said it was put together by their sales and marketing team, not their engineers.
When the City of Dayton dropped their contract, they conceded that because the city had also adopted other violence prevention tactics, they could not measure the effectiveness of the technology at all.
Detroit has a history of experimental policing tactics. From STRESS to broken windows policing, we’ve tried tactics that, when faced with the data later, proved ineffective, and we had to abandon.
This time we need not wait for new data or for a tragedy to unfold.
I don’t want to see the day when a police officer expecting gunshots makes a mistake that costs a life, or someone is imprisoned for a crime they didn’t commit, using bad evidence. These aren’t just hypothetical scenarios – they have happened in other cities.
The chief ends his op-ed admitting there’s no way to isolate ShotSpotter’s effect, but writing that if crime is being reduced, “why mess with it?” There’s an important piece of data, along with the arguments above that may help answer that question — in a study of several cities that ended ShotSpotter, there was no change in their crime metrics from before or after the contract termination.
We should all be proud of the work that we are doing as a city to decrease violent crime. We’re playing our part to contribute to the trend of crime dropping nationwide, and I do believe we are pushing many of the right buttons.
As the chief mentioned, there is a long way to go. Investments in social programs, education, mental health resources and a strong economic foundation for residents are all scientifically supported and effective ways to control crime. Let’s not waste more money on an ineffective and expensive experiment.
Scott Holiday
The writer is a Detroit resident and executive director of Detroit Action, a community-based organization
Submit a letter to the editor at freep.com/letters, and we may publish it online or in print.
This article originally appeared on Detroit Free Press: Detroit police ShotSpotter data don’t tell the whole story | Letters
Reporting by Detroit Free Press / Detroit Free Press
USA TODAY Network via Reuters Connect
By Detroit Free Press | USA TODAY Network
