|
Getting your Trinity Audio player ready...
|
Sharp pops ring out. Acoustic receivers affixed to poles register the sound and send an alert to a command center. There, people employed by the receiver company review the audio before alerting local police. Officers get dispatched to the scene to get eyes on what happened. Simultaneously, or minutes later, 911 calls might come in from residents who heard (or possibly saw the source of) the bangs.
That’s ShotSpotter, a program used by the city from 2011 until June 2026, when the city dropped it from its new budget. In the last five years, it has sent more than 700 alerts to city police. But officers were only able to confirm about 25% of those as shootings or shots-fired incidents, according to New Bedford’s police chief.
That doesn’t mean shots weren’t fired. (Some revolvers don’t leave behind casings and sometimes officers cannot find the evidence.) But amid a budget squeeze, the city cited internal performance metrics to argue the program was no longer worth its cost, which was about $800,000 over the last five years.
Some city councilors have been critical of the decision, saying it will hurt public safety. But the police chief says the program’s accuracy was quite low, so it wasn’t an economical use of resources.
What the data show
Through a public records request, The Light obtained ShotSpotter data from January 2021 to May 2026.
The city did not provide certain data categories, claiming they would not have provided an accurate picture because they were not consistently filled out. In some areas, data cells were blank, so it is unclear what the outcome was.
What is clear is that ShotSpotter alerts have consistently decreased. In 2021, the department received 224 alerts. In 2025, it received 99. Through late May this year, the department received 39.
ShotSpotter totals don’t mirror the frequency of documented shots-fired incidents (including shootings). According to the department, shots-fired incidents were higher in 2023 and 2024 than in 2021, 2022 and 2025. However, some of those confirmed incidents may have happened indoors.
Holly Huntoon, spokesperson for the department, in an email said NBPD “cannot speak as to why the number of alerts may have increased or decreased over time.”
Just under one quarter of alerts over the last five years were verified as shootings based on notes that officers filed after visiting the scene. Police found casings, property damage, victims, blood and CCTV footage, the data reads.
In some cases, the loud boom that receivers detected was a gunshot, sometimes multiple. But other times, the boom that the system flagged as a possible gunshot wasn’t.
According to the data, it was sometimes a firework, an ambulance driving over something that “popped,” a bus tire exploding, cars backfiring, and, in one documented instance, a thunderstorm.
This July 4 at around 7 p.m., the department received a ShotSpotter alert, despite its contract having ended June 30. It turned out to be fireworks.
Police Chief Jason Thody said during a June City Council meeting that at one point, ShotSpotter refunded the city because the program was not operating at agreed-upon standards.
Huntoon said Thody was not available for an interview.
“Despite extensive investigation and a review of any 9-1-1 calls, witness reports, video, search for casings, hospital checks and other evidence, we were unable to confirm 620 (76%) of ShotSpotter alerts,” Huntoon said in an email.
“Each of the 791 alerts were responded to with great urgency, despite the fact that only 24% were ultimately confirmed through a ShotSpotter alert,” Huntoon continued. “That resulted in 620 emergency calls, most often involving multiple officers and further daylight follow-up, where there was no evidence of any shots fired.”
Huntoon said ShotSpotter alerted for 71% of the 241 shooting and shots-fired incidents since 2021 that “met all the criteria to expect an alert.”
The receivers are meant to capture shots outside (not inside), and are not positioned in every part of the city.
Controversial program
SoundThinking, the California-based company that runs ShotSpotter, advertises itself as a security technology company that can detect gunshots in under 60 seconds, allowing a “faster response to save lives.” According to a promotional video from 2023, ShotSpotter was being used by more than 150 cities.
Over the last decade, departments across the country have reconsidered their contracts with the technology.
The Fall River Police Department ditched ShotSpotter in 2018, saying it had missed too many shootings and put out too many false alarms. The then-police chief said it had a 50% accuracy rate. More recently, cities like Chicago and Cambridge have ended or are ending the program.
A 2024 study by the ACLU of Massachusetts found that 70% of alerts by Boston Police Department’s ShotSpotter led to “dead ends.” (Boston PD continues to use it.) In New York City, a 2024 audit concluded the technology was wasting officers’ time, giving false alarms 87% of the time.
Thody presented data to the City Council saying that 911 calls came in for 76% to 90% of shots-fired incidents and shootings, meaning that most of the time, police are getting alerted beyond ShotSpotter.
Based on his personal experience, Lorenzo Gonzalez, NBPD Union president, said a 911 call most often comes in when there is a ShotSpotter activation.
“People in the city are calling to let us know there was just a shooting in this area,” he said.
Gonzalez said ideally he’d like ShotSpotter to stay, since it’s useful for some incidents.
“Is it a perfect system? No. I know it does have its errors and whatnot. That argument could be made about any system, really,” Gonzalez said. “I know that overall, for us, it is beneficial when it does go off.” He said ShotSpotter gave police a general area to search for shell casings or evidence of gunfire. “That’s huge.”
In 2024, U.S. Senators Elizabeth Warren and Ed Markey from Massachusetts wrote to the federal government with concerns about ShotSpotter, saying it may be over-deployed in Black and Latino communities and “perpetuate the over-policing and unjustified surveillance of communities of color.”
A 2011 press release from ShotSpotter announcing its partnership with the city said the receivers would be going in “areas with high incidents of gunfire.”
The city would not disclose where in a four-square-mile section of the city the receivers are located, writing it would “expose the areas not covered by the sensors.”
“As a result, a terrorist or other mal-intentioned individual would be able to maximize the amount of damage exerted on the city, as they would be privy to the locations of the ShotSpotter zones,” the city wrote in response to the records request.
New Bedford is about 20 square miles (11 miles top to bottom and about 2 miles across in some areas). Per SoundThinking, the receivers have a range of one to two miles depending on weather conditions.
Renee Ledbetter, president of the New Bedford chapter of the NAACP and a director of the Shannon Program, which works with youth engaged in or affected by violence, is conflicted.
She sees violence regularly, including kids as young as 12 using guns, and therefore sees the system as a tool that can assist with response times. She also worries about a “bystander mentality”: some residents might hear gunshots but assume a neighbor will call it in, or may be desensitized after hearing shots so often.
But she also recognizes some people have concerns about over-policing in Black and Brown communities.
Ultimately, she says “now is not the time” to get rid of an acoustic tool like ShotSpotter.
“It gives [police] the ability to solve cases. That’s how I’m looking at this because of the violence I see every day,” Ledbetter said.
What does this mean for the state of surveillance?
New Bedford may be jettisoning some “ears,” but its official “eyes” have grown (along with the proliferation of private household door cameras).
Last year, the city installed 20 license plate readers — cameras that capture images of cars and their plates. The camera company, Flock Safety, allows law enforcement agencies to integrate with and access cameras nationwide, enabling police in Texas to search license plate data captured in Massachusetts, for example.
The city operates 249 cameras, according to the department, seven of which are standalone Flock systems, and 13 that are city-owned but equipped with Flock software.
According to NBPD, the Flock system has assisted in several crimes, including shootings.
In December 2024, the city paid Flock $81,000 for the cameras and will have an annual subscription fee of about $36,650, according to the city’s public information officer, Jonathan Darling. Though it’s not an apples to apples comparison (one system is visual, while the other is acoustic and limited to shootings), this newer subscription costs less than ShotSpotter.
Flock systems have gunshot-detection capabilities, but it requires an additional audio detection device, which the city does not use, per Huntoon.
The Light will report on how Flock works, and what the data show with the NBPD’s use of that system, in a forthcoming story. If you have questions, thoughts or concerns about the city’s use of Flock, drop us a line.
Email Anastasia E. Lennon at alennon@newbedfordlight.org.
Editor’s note: Renee Ledbetter is a member of The New Bedford Light’s Advisory Committee. The Light’s newsroom is scrupulously independent. Only the editors decide what to cover and what to publish. Founders, funders and board or advisory members have no influence over editorial content.
