San Diego’s surveillance ordinance is not yet law, but it’s already having an effect on city policy.
The San Diego Police Department said it extended its agreement with ShotSpotter, a gunshot detection company, on a month-to-month rather than annual basis. City staff have the option of keeping the program alive until November 2021, when the terms of the original contract end, but beyond that they’ll need explicit permission from the City Council.
“We didn’t feel it was prudent to lock into a one-year contract knowing that the City Council had concerns about the use of technology,” said Lt. Shawn Takeuchi, an SDPD spokesman.
In November 2020, elected officials approved the first draft of a new law governing the city’s acquisition of devices capable of watching and listening to the public, and laid the groundwork for a privacy advisory board to help oversee those practices. Both are expected to go back to the City Council within the coming months, after undergoing review by the mayor’s office and labor groups.
There’s been growing uneasiness among Black leaders and activists about the way the city has been rolling out technology in public rights of way, and evidence that ShotSpotter, in particular, is not as effective as it’s been portrayed.
For years, police personnel have argued that the audio devices installed around southeastern San Diego reduced violence and connected previously unknown crimes by alerting officers to instances of gunfire without waiting for someone to call 911. One of the intentions was to speed up the city’s emergency response and ensure cops weren’t wasting their time in the process.
The company boasts that it can tell the difference between gunfire and other loud noises — a jackhammer, for instance — and pinpoint the location of those shots to within 25 meters.
Last year, however, a Voice of San Diego review of dispatch records suggested that false alarms were more common than advertised. One officer confided to his colleagues in an email that the cost of the program — now totaling more than $1 million — wasn’t worth it. At a trial in San Francisco several years ago, a forensic analyst for ShotSpotter testified in court that the company’s accuracy guarantee had been invented by its marketing team.
ShotSpotter’s other major selling point is that its technology helps police departments collect ballistics information and DNA evidence from shell casings, which can then be uploaded into law enforcement databases. More often than not, though, cops in San Diego walk away from a ShotSpotter alert without producing a report, because they found nothing at the scene.
Even so, SDPD contends the shell casings it does manage to find are extremely valuable. Capt. Terence Charlot, now an assistant chief, made this point during a presentation to the San Diego public safety committee in October 2017.
“The crime lab has extracted a useable DNA profile in 56 percent of the cases analyzed,” he said.
Yet the city’s own records suggest that that figure wasn’t quite right at the time, and isn’t right today.
When pressed for more information, SDPD noted that there’d been a turnover in staff at the crime lab in recent years. Instead, Takeuchi analyzed the data again last week and found that, at the time of the original presentation, the crime lab had managed to extract a useable DNA profile 47 percent of the time. But if you break out the data even further — by extending it through the length of the program and looking at the individual casings recovered, rather than the batches collected from each ShotSpotter alert — you get a different result.
Separately, VOSD reviewed almost four years’ worth of data and found that the crime lab has managed to extract DNA evidence from only 30 percent of the individual casings. About two-thirds of the time, the casing contained no DNA evidence or not enough to build a profile.
That 30 percent figure is not insignificant. DNA evidence is difficult to pull from shell casings generally, and SDPD has helped pioneer a more reliable testing technique in the United States. But that 30 percent figure is still below what police personnel originally offered as a justification for keeping the program going.
Takeuchi’s analysis also found that SDPD’s crime lab, roughly a year into the ShotSpotter contract, was getting either a useable DNA profile or useable ballistics information that could identify the type of gun involved more than 70 percent of the time. Charlot had mentioned this stat at his 2017 presentation as well, but VOSD couldn’t independently verify it.
The overarching issue is that, besides the occasional update, there’s been no meaningful vetting of the audio surveillance system, which was placed in predominantly Black neighborhoods, because the original agreement with ShotSpotter never went through the City Council. It was signed in 2016 and executed with seized asset funds, mostly coming from the San Diego County District Attorney’s Office.
At the time, officials were talking about using ShotSpotter as part of a regional task force dedicated to tracking and preventing gun-related crimes. Emails show that county prosecutors weren’t the only ones pushing the project — the U.S. Department of Justice was involved in early discussions.
Through it all, one of the biggest proponents at the city level was Jennifer Shen, then the head of SDPD’s crime lab. In early 2015, almost two years before the devices were installed, she sent a message to another manager and asked: “What is ShotSpotter, and why does Denver have it and we don’t?”
Shen went on to highlight the technology in a speech she gave to the San Diego Police Foundation in 2018. She left the city in late 2019, following a VOSD investigation showing she used a less rigorous testing standard in order to clear the crime lab’s significant rape kit backlog, and did not return a request for comment for this story.
Shelley Zimmerman, the former police chief, was also a big supporter of the technology at the time. In 2016, she sent Shen and a pair of her underlings, including current police chief David Nisleit, to Milwaukee to see for themselves how that city was incorporating ShotSpotter into its investigative work.
“Although our gun crime numbers do not match yours, we feel that the implementation of ShotSpotter, along with the intelligence generated from analysis of collected cartridge cases, will enable us to reduce gun violence in the high crime area that will be covered by this new program,” she wrote to Milwaukee’s police chief.
The city only agreed to test out the technology for a year. ShotSpotter gave officials the option to extend the contract for four additional years. SDPD has taken three of those extensions and said it agreed in the final year of the contract to switch to a monthly renewal.
In other words, what began as a pilot program, championed and approved unilaterally by law enforcement agencies, has evolved into a nearly five-year initiative with mixed results. And it’s not just San Diego. In 2016, Reveal found that cities across the country had few tangible results despite spending millions on the technology and responding to thousands of alerts.
To this day, the precise locations of the devices, although attached to public infrastructure, remain a secret in San Diego. Officials have declined to release this information, citing public safety.
SDPD, in the meantime, has said that it intends to review all relevant crime statistics and work with elected officials to see if ShotSpotter is still in the community’s best interest.
Of course, not everyone is on board, and they’ve made their opinions known from the start.
Before she was elected, Councilwoman Monica Montgomery Steppe appeared at the 2017 public safety committee meeting — a body that she now chairs — to warn that SDPD was doing harm to its reputation in Black communities. She argued that the city’s stats about DNA collection and response times didn’t justify its use.
“Even if this were the greatest thing in the world, which it is not, to come here and try to say that it is, is misleading,” she said. “It hurts us all because it leaves us farther away from establishing that accountability and trust that we need.”
As the meeting progressed, officials teased out one of the most underappreciated aspects of surveillance generally. Devices like ShotSpotter’s are supposed to have a psychological effect on communities, reminding anyone who comes into the orbit that they’re being monitored for potentially criminal behavior — in this case, for loud noises that might signal a gunshot.
During her questioning of police leaders, then-Councilwoman Lorie Zapf argued that cameras in Ocean Beach, along with signs telling people they’re under surveillance, had helped reduce crime. She wondered if deterrence was the goal with ShotSpotter too.
“Very much so,” Zimmerman responded. “We say that we would prefer to prevent crime instead of responding to crime.”
Kara Grant contributed to this report.