Research at the University of Michigan finds law enforcement technology aimed at detecting gunshot sounds and notifying police in close to real time raises serious concerns about accuracy, effectiveness, cost and systemic bias.
The policy brief calls into question claims by ShotSpotter, the most prevalent company employing acoustic gun detection systems, that the technology improves safety, reduces crime and promotes positive relationships between communities and police. The brief, Acoustic Gunshot Detection Systems: Community & Policy Considerations, also offers remedies for communities concerned about the technology's deployment.
"The performance equation ShotSpotter uses only counts errors based on failure to identify a verified gunshot, or a mislocated verified gunshot," writes Jillian Mammino, a second year Master of Public Policy student at the Ford School of Public Policy. "Meanwhile, the company claims ShotSpotter has only a 0.5% false positive rate, yet independent studies and customer reviews highlight false positives as the number one operational concern."
Mammino, who wrote the brief as part of the Science, Technology and Public Policy program's community partnerships work, also evaluates the broader claims about community engagement and improved policing, noting evidence the system "actively reduces police/citizen collaboration."
Citing studies in St. Louis and Chicago, she notes the St. Louis Police Department had a decrease in civilian reported gunfire calls in the areas covered. In Chicago, 89% of ShotSpotter calls turned up no gun-related crime, and 86% led to no report of any crime at all.
The "false alarms," the study says, mean the actual cost of ShotSpotter to Chicago is much higher than the $10 million a year annual subscription due to the unnecessary deployment and use of limited police resources. They also likely contribute to higher policing costs in other cities.
Finally, she said that as with other types of technology deployed by law enforcement, including facial recognition and pre-trial algorithmic risk assessments, ShotSpotter reproduces "systemic biases by relying on past data that has been shaped by over policing, biased enforcement patterns, and ongoing disinvestment in marginalized communities."
The Science, Technology and Public Policy program works with community organizations that have concerns related to a current or anticipated science or technology issue, and provides support to help them engage in technical and policy discussions. Through its Community Partnerships Initiative, the program shares information with community organizations interested in learning more about technology that has or may affect their members.
With this information, We the People Michigan, an organization that believes in empowering communities to determine their own futures, created material to use in its upcoming campaign against expanding a ShotSpotter contract in Detroit.
Communities concerned with local law enforcement's purchase and deployment of the technology can advocate for various policy responses including a ban, moratorium, community oversight or technology assessment and deliberative democracy, in which community knowledge and expertise informs the consideration of a technology's benefits and risks.
"Given the problems with accuracy, efficacy, cost, and bias described above, it seems likely that the funds cities allocate to ShotSpotter would be better spent on community investments that are proven to reduce crime, such as after-school programs, drug treatment programs, and poverty alleviation," the study says.