Credit: Madison Muller

Correction, 8/3/2022: This story has been updated to remove references to an analysis of the locations of ShotSpotter surveillance devices conducted by Lucy Parsons Labs. The analysis relied on work permits filed by Motorola Solutions and the Office of Emergency Management and Communications as well as data from the Department of Transportation, and incorrectly identified hundreds of devices as ShotSpotter equipment that were not. After Lucy Parsons Labs alerted the Reader about flaws in the analysis, we removed all references to locations of ShotSpotter devices in the article, and replaced a map of ShotSpotter devices we created based on the original analysis with one published by the Office of the Inspector General in 2021.

Before last March, you might not have heard of ShotSpotter. That month, news of 13-year-old Adam Toledo’s killing by a police officer in Little Village rang through Chicago just as resoundingly as the alleged noise of gunshots that brought cops to his location in the first place.  

Almost as soon as the news of Toledo’s death broke, activists began raising questions about the private gunshot-detection system that summoned police to the scene. That summer, the activists’ voices grew even louder when it became public that the city had quietly extended its $33 million, three-year contract with ShotSpotter by another two years, through August 2023. 

ShotSpotter markets its technology as a “proactive” tool that hears gunshots and gets police to potential crime scenes faster than 911 calls. 

But sound is a tricky thing. It travels, echoes, and reverberates, and can be muffled, distorted, or unclear. How can ShotSpotter sensors tell the difference between a gunshot or firework? They apparently can’t—at least not as accurately as the company has publicly asserted. And there is little evidence to suggest that the Chicago police (CPD) or ShotSpotter test the devices to see how they register different loud noises once deployed. 

ShotSpotter’s primary purpose is to hear gunshots, but according to a report published by the city’s Office of the Inspector General (OIG) last August, only 9.1 percent of alerts generated between January 1, 2020, and May 31, 2021, actually resulted in police finding evidence of a gun crime. ​And a 2011 study commissioned by the company found that trucks, motorcycles, helicopters, fireworks, construction, trash pickup, and church bells, among other sounds, have all triggered false positive alerts, mistaking these sounds for gunshots.

In its 2021 report, the OIG found that the very presence of ShotSpotter changed the behavior of police and how they worked in areas where ShotSpotter devices were present. The report found that CPD officers’ “generalized perceptions of the frequency of ShotSpotter alerts in a given area may be substantively changing policing behavior.”

Where ShotSpotter devices are has paradoxically been a closely guarded secret: the CPD won’t acknowledge the exact locations of the devices, despite the fact that they can be seen from the street. The department steadfastly denies Freedom of Information Act (FOIA) requests for ShotSpotter locations by claiming that such information is exempt from disclosure because the metadata of the device locations may include private addresses. In response to one FOIA request, the department claimed that because ShotSpotter is used for law enforcement, “the release of such information could endanger the life and safety of property owners who are cooperating with police.” 

The OIG’s report revealed that the city’s violence reduction dashboard confirmed that as of May 2021, ShotSpotter devices had been installed in 12 police districts. (The OIG noted that 0.6 percent of alerts were located in a district outside the 12 with confirmed ShotSpotter sensors.) 

The devices are concentrated in predominantly Black and Brown neighborhoods on the south and west sides. 

That disparity raises concerns that ShotSpotter disproportionately increases the likelihood of more—and more aggressive—policing in communities that are already overpoliced. According to a study by the MacArthur Justice Center, even residents in the vicinity of a ShotSpotter device can be threatened by overpolicing and “volatile deployments.”  

“ShotSpotter primes police to believe that they are heading to a dangerous location where a person has just fired a gun,” the study says. “Any resident who happens to be in the vicinity of a ShotSpotter alert will be a target of police suspicion or worse.”

According to the OIG’s report, there is also a pattern of police stopping people who happen to be in the vicinity of ShotSpotter and using the device’s proximity to rationalize the stops and subsequent pat-downs. 

In response to detailed questions from the Reader about ShotSpotter, a CPD spokesperson provided a statement that said in part, “[t]he detection technology is among a host of tools used by the Chicago Police Department to keep the public safe and ultimately save lives.” 

Adam Toledo’s death wasn’t the only time the devices made headlines. In March, the Associated Press reported that Michael Williams had been jailed in 2020 for supposedly killing a young man from his neighborhood. During the citywide unrest that followed the police murder of George Floyd in Minneapolis, Williams gave a ride to a man who was later killed. Williams was arrested for the killing based on evidence that prosecutors later said was unsubstantial—a “noiseless security video showing a car driving through an intersection, and a loud bang picked up by a network of surveillance microphones.” 

Williams’s case was dismissed and he was later released from Cook County Jail in July of 2021, but his arrest and the time he spent in the jail—during which he contracted COVID-19 twice—left him “shaken” and feeling unsafe when he walks around his own community. 

Another key aspect of Williams’s case is that there appeared to be modifications of the data after the event occurred, which is not unique to this case. According to Vice, “months later and after ‘post-processing,’ another ShotSpotter analyst changed the alert’s coordinates to a location on South Stony Island Drive near where Williams’ car was seen on camera.”

Update 5/23/22: After this story was published, a ShotSpotter representative contacted the Reader to note that the technician had not changed the latitudinal and longitudinal coordinates of the alert, which they said remained consistent. A forensic report that the representative provided to the Reader stated in its conclusion, “ShotSpotter recorded the incident . . . and located it at 5700 S. Lake Shore Dr. After post-process analysis, the incident is found to have occurred at or near 63rd Street & Stony Island.”

According to the OIG’s report, ShotSpotter devices located near the edge of one district can occasionally pick up sounds in another. Jonathan Manes, an attorney at MacArthur Justice Center who leads MacArthur’s Illinois office on surveillance technology, said this is to be expected because neither ShotSpotter or CPD tests the devices to determine how they respond to different types of sound.

While ShotSpotter claims to have an accuracy rate of 97 percent, Manes said that misrepresents the facts. 

“They continue to repeat every chance they get that their system is supposedly 97 percent accurate,” he said. “It’s not, and that number is not an accurate number.” The figure is based on the assumption that every alert ShotSpotter creates is based on an actual gunshot, unless the police department disagrees, according to Manes. “We know that Chicago police never send a complaint for false alerts,” he said. “That’s just not an accuracy number, it’s a tally of customer complaints, and the fact that they keep calling it an accuracy number is, in my view, just dishonest and misleading.”

ShotSpotter’s supporters and critics don’t shy away from publicly voicing their thoughts on the devices, and Manes is no exception. 

He regularly reports his findings on ShotSpotter’s inaccuracy and ineffectiveness on Twitter, and about a year ago, a Twitter user under the username of @slayercapital, an account that has since been deleted, started responding, calling his claims against ShotSpotter “disgusting and despicable.” 

The account’s owner claimed to have no involvement with ShotSpotter, but Manes said he believes the account was in fact run by Doug McFarlin, ShotSpotter’s vice president. Before the account was anonymized, it included McFarlin’s first name, and most of the account’s followers are ShotSpotter executives. 

That wasn’t the only instance where ShotSpotter’s C-suite attempted to use Twitter to further their agenda. Last August, when the anonymous account criticized Action Center on Race and the Economy organizer Alyx Goodwin’s takes on ShotSpotter, she took to Twitter. She says she tweeted questions about whether the anonymous messages were being fired out from a burner account run by ShotSpotter CEO Ralph Clark.

Clark responded from his account to say the anonymous account wasn’t him. After Goodwin accused Clark of overpolicing Black and Brown communities, she said he called the accusations false. 

Goodwin said she also noted an apparent shift in marketing from ShotSpotter. While the surveillance company often boasted its 97 percent accuracy rate, last summer, Goodwin said she saw the company shift to focus on response times, and the ability to get police to scenes quickly. 

“Their claim to fame, that I think they’re trying to really market, is that they cut down police response times,” Goodwin said. 

Manes concurred. He said that the company’s marketing changes over time with what they believe “resonates in the moment.”

“When discussions about policing were focused on community policing, and having officers sort of walking the beat and that sort of thing, they were marketing their product as a good way to get officers into the community,” Manes said.   

“Now, I don’t think that sending officers chasing down nonexistent gunfire is an especially conducive way to improve police-community relations,” Manes added, “but that was the pitch they made then.”

A guide to midwestern radiators and their calls

In the 2015 horror film The Blackcoat’s Daughter, there’s a midnight scene where a teenager named Kat slips out of her room and into her boarding school’s dorm basement. An older girl named Rose, apparently the only other student left at the school over winter break, grows curious and follows. Old pipes creak and whine…