In the first eleven days of 2019, six people were shot and killed in Durham. There were two double homicides, including one whose victims were a young mother and her ten-month-old baby. Police consider three deadly incidents cases of domestic violence. One man was found dead after a standoff with county deputies, his son charged with his murder. And, in broad daylight, two teenage girls were struck by bullets fired by gang members shooting at each other from their vehicles, inflicting non-life-threatening injuries.
“Every time a person is shot in Durham, it rips a hole in a family. It rips a hole in a neighborhood. It rips a hole in a community,” Durham Mayor Steve Schewel said at a news conference last week.
There were fewer gun crimes in 2018 than the three years prior, despite Durham’s population growth. But these recent shootings have alarmed the community, kept officers working long hours, and caused elected officials to examine the city’s approach to combating gun violence.
It’s against this backdrop that Durham officials, police, and sheriff’s office representatives will get a pitch next week from a company called ShotSpotter to install its gunshot-detection technology around the city. If the tech is installed, sensors placed throughout the city will listen for “short, explosive” sounds like gunshots. When the sensors are activated, an audio clip will be sent to machine, and in some cases human, reviewers to confirm whether the recording is actually of gunfire. If it is, local police will be dispatched to the scene. According to ShotSpotter, this all takes place within seconds.
The goal is to get police to the scene faster, dispatch them to incidents that aren’t reported to 911, and give them information on the kind of evidence—like the number of shots and caliber of the gun—to look for when they arrive. ShotSpotter boasts impressive results, saying the devices “detect 90%+ of gunfire incidents with a precise location in less than sixty seconds.”
But questions have been raised over whether ShotSpotter actually helps reduce crime—and whether the sensors could be used to record more than rounds fired in the street.
Developed in the 1990s from earthquake-detection technology, ShotSpotter is in use in about ninety cities across the U.S., including Rocky Mount, Wilmington, Goldsboro, and Greenville in North Carolina.
According to company spokeswoman Liz Einbinder, ShotSpotter sensors can detect gunshots up to a mile away. The devices don’t have live audio streaming capabilities and are only activated when a gunshot-like sound is detected. The three closest sensors then “triangulate the exact location of the event within eighty-two feet,” according to the company. The sensors—ideally installed thirty feet above the ground, most commonly on the tops of public and commercial buildings—can’t detect gunfire indoors.
City council member Mark-Anthony Middleton has been advocating for the city to consider ShotSpotter since 2015, when he saw it in action as part of a local delegation that traveled to Boston. Middleton grew up in Brooklyn and remembers gunfire being so commonplace that neighbors stopped calling the cops.
“Sometimes,” he says, “what would happen is that when you get up the next morning, there’s a body in the park because somebody actually did get shot, but because we had become so desensitized, people had stopped calling the police.”
While he wants to hear more from ShotSpotter and the community, he says that, given the level of gun violence in Durham, it would be irresponsible not to discuss the technology.
City officials first considered installing ShotSpotter nearly two years ago but didn’t move forward because they worried about the expense. As city manager Thomas Bonfield explains, gun crimes aren’t concentrated in one neighborhood, which means the city would need to install sensors across the city to be effective. The larger the coverage area, the higher the bill.
In 2017, ShotSpotter crafted a pricing proposal covering parts of East Durham and areas around N.C. Central. According to the proposal, the city would pay a one-time $10,000 “onboarding” fee, $10,000 per square mile for installation, and an annual service fee of $65,000 per square mile. To cover the three-square-mile area in East Durham outlined by the proposal, Durham would pay $235,000 for the first year.
The city has been talking with N.C. Central and Duke about splitting the cost, which wouldn’t be unique. The city of Greenville is sharing its costs with East Carolina University, the local medical center, the sheriff’s office, and the public housing authority.
“It’s expensive, but so are human lives,” Middleton says. “What price do you place on a Bull City resident’s life?”
While working for the Atlanta Police Department, Durham Police Chief C.J. Davis “worked with the ShotSpotter representatives in Atlanta to assess its use,” according to the Durham Police Department. Since her departure in 2016, Atlanta has implemented ShotSpotter. (Davis declined an interview request.)
Other cities Durham consulted say the technology helped officers respond to more shootings faster, Bonfield says. But it hasn’t always reduced the number of gun crimes.
“Everybody has confirmed that the system was very, very accurate,” Bonfield says. “There’s no question about the accuracy. The question was, what’s the result of getting there quicker?”
Sergeant Brad Summerlin of the Rocky Mount Police Department says shots-fired incidents there have decreased by 60 percent since 2012, the first full year after the city became the first in North Carolina to adopt ShotSpotter.
According to Summerlin, the technology is helping officers get guns off the street and arrest the people using them illegally, leading to a decline in gun crimes. While the sensors have picked up firecrackers and other loud noises, he says, those alerts get filtered out and, to his knowledge, officers had never been dispatched to a false alarm.
That hasn’t been the experience of all cities employing the technology.
In 2016, Forbes obtained ShotSpotter data and found that, across seven cities, authorities dispatched by a ShotSpotter alert were unable to find evidence of gunshots when they arrived between 30 and 70 percent of the time.
“The majority of ShotSpotter alerts lead to police closing the incident with words such as ‘unfounded,’ ‘unable to locate,’ or ‘gone on arrival,’ law enforcement jargon for: ‘We didn’t find anything,’” the story says. (Outside of annual highlights, ShotSpotter doesn’t share its data.) Over the course of about a year and a half, Forbes reported, ShotSpotter devices in Wilmington produced 1,278 results. A third of the time, law enforcement didn’t find evidence of a shooting, and only five arrests were made.
In 2016, Charlotte decided ShotSpotter hadn’t led to enough arrests to justify renewing an annual $160,000 contract, The Charlotte Observer reported.
It’s unclear why some officers dispatched by ShotSpotter alerts don’t find evidence of a shooting. But it is clear that ShotSpotter, by design, sends cops to the neighborhoods being monitored by its sensors more often than they were before—and when they arrive, they’re prepared to find a shooter. That could create tension in communities that are already heavily policed, says state ACLU spokeswoman Molly Rivera.
“We know that ShotSpotter, just like other surveillance technologies that are used by police, are used especially against communities of color, and it creates an oppressive and stigmatizing environment in which that community and the people who live within that community are treated like prospective criminals,” says Rivera. “They’re going in expecting to find an active shooter, and entering a community like that creates a particular environment.”
Critics have also suggested that ShotSpotter could violate people’s Fourth Amendment rights against unreasonable searches by recording dialogue, intentionally or not. While ShotSpotter sensors aren’t activated by human voices, they have recorded people talking around gunfire—including a few cases in which a victim’s dying words have been captured and used to secure a conviction.
“Our technology and privacy policy allow us to provide only a few seconds of audio just before and after a gunshot,” says Einbinder. “If voices are concurrent with the gunshot, and they are loud enough and close enough to a sensor, then they may be captured in the audio alert snippet. But that is a very rare occurrence and far from a conversation.”
Jeff Welty, an associate professor of public law and government at UNC, says that while “I don’t think the question is completely settled,” it’s unlikely that the technology would rise to the level of a Fourth Amendment violation because these conversations “are typically in outdoor public locations” where courts may find there’s no “reasonable expectation of privacy.”
“While there might not be specific privacy concerns with the technology itself, the real question is, well, what else can ShotSpotter do and how else can it be used?” asks state ACLU communications director Mike Meno. “That raises questions like: Are there going to be audio recordings? Are there going to be video recordings? What would trigger those recordings,? Where in the city is it going to be used?”
Some cities have synced devices to activate video cameras when gunshots are detected. At least one, Boston, linked the sensors to ankle monitoring devices, to check whether parolees are close to gunshots and might be violating their parole conditions. In October, ShotSpotter acquired the predictive policing tool HunchLab, which is offered as an add-on. (Bonfield says the city isn’t currently considering syncing surveillance cameras with ShotSpotter. DPD analyst Jason Shiess says the department already employs several tools to “forecast” where crimes may occur, and has for more than a decade.)
Meno says if the city implements ShotSpotter, there needs to be a public process to determine where the sensors would be placed, how they would be used, and what policies would safeguard the technology from abuse.
While ShotSpotter has come up in city council meetings, the council hasn’t yet discussed implementation, which it would have to approve via the annual budget. City and county leaders emphasized during last week’s press conference that Durham won’t eradicate gun violence without addressing its root causes, such as unemployment, and the ability to enact local gun laws, which state law preempts.
Middleton says he knows ShotSpotter won’t be a panacea, but the city should consider every tool at its disposal.
“Crime overall in Durham is down, but when a family member of yours has been shot, that’s not really a compelling argument to you,” he says. “If there’s a tool we can avail ourselves of that can possibly save one life, I don’t see how we cannot have a conversation about it.”
Contact staff writer Sarah Willets by email at swillets@indyweek.com, by phone at 919-286-1972, or on Twitter @sarah_willets.
