Christina Wilcox hears the blast of gunfire echo through her Shawnee neighborhood a few times a week.
She’s learned to quickly assess the threat: Are her kids safe? Doors locked? Should she drop to the ground?
Not a consideration: calling the police.
“I don’t do nothing wrong and I’m going to keep it like that,” she said. “I don’t want them coming to my door and someone down the street sees and thinks I’m saying something.”
This unwillingness to alert police to shootings is not uncommon, and police know it. That’s why in 2017 the Louisville Metro Police Department installed ShotSpotter, an array of microphones across six square miles of the city designed to listen for gunfire and automatically notify officers.
But a Kentucky Center for Investigative Reporting analysis of police data shows the system provides no guarantee police will stay on the scene very long — or actually investigate.
Officers conduct required surveys of the area on fewer than two out of every 10 ShotSpotter calls, according to data from the police department and the city’s MetroSafe 911 service. The calls rarely lead to an arrest.
The officers typically clear scenes within minutes, and they seldom find evidence of a shooting, according to police data.
As Louisville Mayor Greg Fischer looks to free up $35 million to pay for the city’s growing pension obligation, he is threatening to get rid of ShotSpotter. The service costs $390,000 for technical assistance, system maintenance and data storage, according to the city’s contract with the San Francisco-based company. To date, the city has paid more than $1.2 million for the service.
Louisville Metro Police Chief Steve Conrad said recently that, combined with other proposed police cuts, losing ShotSpotter would cause “bleeding” and contribute to a “slow train wreck” for public safety in Louisville.
LMPD spokesperson Jessie Halladay said the arrests ShotSpotter brings are a crime deterrent and a lot of the gunshots it alerts police to aren't reported. She speculated that more officers probably do surveys than the data shows, but they forget to report them.
But some experts question whether the lack of tangible benefits make the service worth the cost.
Police Praise Community Connection, But Rarely Stay Long
LMPD officials won’t say exactly where the equipment is located, citing public safety concerns. But data shows ShotSpotter has detected gunshots across Russell and Shawnee and in portions of Portland, Chickasaw, Smoketown and Old Louisville.
Officers were dispatched to about 3,700 unique gunshot incidents between June 2017 and this March, according to ShotSpotter data provided by LMPD.
That data excludes repeat calls when gunshots are heard within 3 minutes and 500 feet of each other, and calls on holidays when fireworks — and false positives — are rampant.
Police department policy requires officers responding to ShotSpotter alerts to treat the incident as a crime in progress, just like a holdup alarm or a violent crime. Officers are supposed to survey the neighborhood and canvass at least two houses in either direction of the alert location, which ShotSpotter technicians provide as a specific address.
But police data shows that in 2017 and 2018, officers did not survey the neighborhood for 85% of the incidents.
Wilcox works overnights at a nursing home, and one morning a few months ago, Wilcox heard a volley of shots as she arrived home.
She waited in her car. Within minutes she saw police officers in the alley behind her home. She didn’t see them talk to any residents or knock on any doors.
And within five or 10 minutes, Wilcox said, they were gone.
City data shows Wilcox’s block on South 42nd Street has had more ShotSpotter notifications than any other block in the city from June 2017 to April of this year — 62 reported gunshots after holiday calls are removed.
When officers were dispatched to Wilcox’s block, the median time they spent was 12 minutes before they left again.
Lt. James Cirillo, who oversees the department’s use of ShotSpotter, said officers might not conduct a survey for many reasons. It could be late at night, he said, and the officers don’t want to disturb residents. Or officers may be inundated with other waiting calls.
Even when officers do canvass an area, bullet shell casings can be tough to find at night and in tall, overgrown grass, he said. Shooters sometimes pick up casings, or they shoot from inside a car and drive away.
“If the runs are coming in, you know, we can't spend all day on something where we're not really finding anybody injured, any property damage or anything like that,” Cirillo said.
Police Know About More Gunshots But Make Few Arrests
LMPD’s publicly released ShotSpotter data doesn’t include metrics on how many people are shot, but reported gunshots vastly outnumber actual shootings. The police reported 323 shootings in which someone was injured or killed across the whole city in 2018, while more than 2,000 gunshot incidents were reported in the 6 square miles of the ShotSpotter’s range.
Earlier this month, Cirillo said ShotSpotter has led officers to 13 gunshot victims this year. And police data shows officers have recovered 13 guns on ShotSpotter calls.
But, 89% of the time, according to police data, their canvass didn’t turn up any bullet casings.
Suspects rarely stay on the scene after shooting at someone, Cirillo said. Witnesses can be reluctant to share information with police.
Ultimately, few of the ShotSpotter calls result in arrests. Since ShotSpotter was installed in 2017 until late March, officers have made 58 arrests after responding to ShotSpotter alerts.
Matt Stroud, a Pittsburgh-based journalist and author of a book, “Thin Blue Lie,” examining police use of developing technologies, said based on the number of arrests LMPD makes related to ShotSpotter and how long they spend on a ShotSpotter call, it doesn’t seem worth the money.
“Why would police leaders in Louisville continue to spend money on a technology with a success rate of about one-tenth of one percent?” Stroud said.
Records show all the arrests are not related to gun crimes.
More than a third of 2019 arrests reviewed by KyCIR did not include a charge related to guns. One of the arrests was on Christina Wilcox’s street, where three teens were arrested after one allegedly shot a pistol on the back deck of a home, according to a police report.
All three were charged with possession of marijuana and drug paraphernalia, but no gun offenses.
While Cirillo and LMPD’s chief downplay the priority on arrests related to ShotSpotter, they praise the service for its investigative and community relationship-building abilities.
Within minutes, officers can be en route to a scene, regardless of whether a resident called 911. Cirillo said for nearly 80 percent of ShotSpotter alerts, there is no corresponding 911 call, so they would have never learned about those gunshots.
This means police officers are tasked with responding to thousands of more calls than they otherwise would. Cirillo admits it can be burdensome and frustrating, especially when officers don’t turn up any evidence.
But the benefit, he said, is the perceived community trust that comes with the added presence of police.
“You gain some support from the community when they see police are coming out and at least checking on the area,” he said.
But building relationships with community members depends more on interaction than just showing up, said Eric Piza, an associate professor at the John Jay College of Criminal Justice in New York City.
For this, Piza said officers need to show up, get out of their car, walk around and talk with people.
People living on South 42nd Street say they rarely see this.
Faye Young lives a few doors down from Wilcox. He’s a security officer at a local hospital.
He’s lived in his brick bungalow house, with a shaded porch and trim lawn, for more than a decade. He’s heard plenty of shots, but the police have never knocked on his door afterwards, he said. He’s never gotten a door hanger that LMPD policy requires officers to distribute to let them know a gunshot was reported.
“Police aren’t always coming around here,” he said. “I don’t like that.”
‘I don’t think you can accomplish that much in 10 minutes’
While police do sometimes spend hours on the scene of a ShotSpotter alert from the time they’re dispatched to when they clear the call, the median duration of time is about 10 minutes.
This raises questions about how effective officers are in building trust and relationships with residents, said Samuel Walker, a policing expert and professor emeritus in the School of Criminology and Criminal Justice at the University of Nebraska at Omaha.
“Given this includes the travel time to get there, I don’t think you can accomplish that much in ten minutes,” he said.
Nicolai Jilek, the president of the River City Fraternal Order of Police, said there are many reasons why officers may be spending little time on a scene -- officers may get additional information about the shooting that sends them elsewhere, or a more pressing call may come across the radio, or perhaps no witnesses want to talk to them.
“Our officers want to do right, they want to solve crimes,” Jilek said. “I have no doubt that they put in every bit of effort to find something, because if they find something, that's a good day.”
And if the department loses more officers, which is another possibility as the city deals with its budget woes, the time officers can spend on a scene will likely drop even lower as they’re forced to “just go from call to call, run to run,” Jilek said.
“It kind of makes it difficult for officers to have the time to be there long enough to develop those relationships,” he said
Because of this, he said ShotSpotter is a “complicated issue.”
“There are definite benefits,” he said. “But the value, I guess, is perhaps debatable.”
Fischer, Louisville’s mayor, said cutting service to ShotSpotter is not something he wants to do.
Fischer said in a press conference earlier this month that cutting ShotSpotter can free up funds to pay for five police officers, or cover the costs associated with community center operations, immunization clinics and more.
“These now are the kinds of decisions that we are being forced to make,” he said.
The residents along South 42nd Street will be paying close attention to these decisions. They’re aware of the crunch, and what they might lose.
For Wilcox, scrapping ShotSpotter seems counterproductive. She likes the sense of security that comes with knowing that someone — or something — knows what’s happening on her street.
Savings should come from the top, she said, like the Mayor’s office. Not the bottom, she said, from neighborhoods like her own.
Correction: The title of the book, "Thin Blue Lie," was incorrect in a previous version of this story. Reporter Jacob Ryan can be reached at jryan@kycir.org and (502) 814.6559.