The city of Chicago is on edge, haunted by its past and fearful of what lies ahead, for once again a police officer has killed a child on its streets.
On March 29 in Little Village, a predominantly Latino neighborhood on the West Side, police pursued and fatally shot 13-year-old Adam Toledo in what the police department has described as an “armed confrontation.”
In the days since, details have trickled out in a manner that has done little to dispel the climate of distrust that now attends police shootings. On the contrary, the incident has reawakened the collective civic trauma inflicted by the 2014 police murder of 17-year-old Laquan McDonald.
In contrast to the administration of her predecessor Rahm Emanuel, which withheld dashcam video of McDonald’s murder from the public for 13 months and then suffered an irremediable collapse of credibility when it was finally released, Mayor Lori Lightfoot’s administration has said it will make relevant video footage public as soon as the Toledo family has had an opportunity to view it.
In anticipation of protests and possible civil unrest after the video footage is released, the Chicago Police Department has informed officers that days off will be canceled and that they will move to 12-hour shifts.
At an April 5 news conference, Chicago Police Superintendent David Brown said that “one of his greatest fears,” now realized, “has been a deadly encounter between one of our officers and a juvenile.” He then gave voice to the inevitable refrain on such occasions: “Our officers must make split-second decisions when it comes to the use of deadly force, and that is a heavy burden.”
Brown described the incident. “At approximately 2:36 a.m.,” he said, “ShotSpotter detected eight gunshots” at a particular location in Little Village. ShotSpotter, he explained, “is a gun detection system that operates through a series of sensors to identify potential gunshots” and “alerts officers in real time to the location of gunfire.”
Officers received the notification at 2:37 a.m. and reached the scene in less than a minute. When they arrived, Brown recounted, “they observed two males in a nearby alley. Both males fled. One was armed with a handgun. A foot pursuit ensued, which resulted in a confrontation in the alley.” An officer shot once, fatally striking Toledo in the chest. “A gun was recovered.”
In her remarks at the news conference, Lightfoot spoke forcefully of the urgent need for CPD to develop a new policy on foot pursuits. Although there is much we don’t yet know about the incident, she said, “one issue that is clear is that a foot chase was involved.”
The mayor went on to describe the problem. “Foot pursuits present a significant safety issue, officer safety, but also community safety for the pursued and bystanders,” she said. “And CPD engages in hundreds of foot pursuits annually and many every single day. Police get a call, they see a potential suspect, their adrenaline is pumping, and oftentimes they get separated from their partner, so they’re running on their own through a dense, often dark, urban environment. And to add to that the person being pursued often has a firearm or is suspected to and so does the officer.”
This combination of factors “creates a dangerous environment for all involved — the officer, the person being pursued, and any bystanders,” she said. “So now we cannot and will not push the foot pursuit policy reform off for another day.”
She might have added that foot pursuits most often take place in Black and Latino neighborhoods and that the frantic, adrenaline-saturated dynamics she evoked generate a mode of attention in which threat assessment is likely to be shaped by implicit bias. At such moments, “split-second decision” is a misnomer, for it is the sheer unthinking momentum of the interaction rather than a deliberate decision that results in the use of deadly force.
Does ShotSpotter Increase Danger of Police Shootings?
A good place to begin the reform process Lightfoot has called for would be a critical assessment of ShotSpotter. An ongoing investigation by the MacArthur Justice Center of Northwestern University Law School has yielded evidence that this high-tech tool is wasteful, alienating for community members, and generates intolerable risks of avoidable harms.
Once we know more about what happened on March 29, it may well prove to be the case that ShotSpotter worked as intended. That makes the Toledo incident a powerful occasion for assessing the hypothesis that this technology, as used by CPD, creates an unacceptable risk of producing “split-second” situations — situations that would not otherwise occur — in which officers respond to perceived threats with deadly force.
On their website, ShotSpotter claims 97 percent accuracy. That figure is not, however, the result of rigorous research. In a 2017, a ShotSpotter forensic analyst testified in an attempted murder case in San Francisco. When asked about the company’s guarantee of accuracy, he stated, “Our guarantee was put together by our sales and marketing department, not our engineers.”
Gunshot detection systems increase demands for police resources but do not result in reductions in violent crimes or increases in the number of confirmed shootings.
Remarkably, there are no independent, peer-reviewed studies of ShotSpotter efficacy. There are, however, two prominent studies that conclude gunshot detection systems increase demands for police resources but do not result in reductions in violent crimes or increases in the number of confirmed shootings. (Both studies were published in the Journal of Experimental Criminology; one focused on ShotSpotter in St. Louis, the other on a comparable gunshot detection system in Philadelphia.) For these reasons, a number of cities, San Antonio and Charlotte among them, have canceled their contracts with ShotSpotter.
The team at the MacArthur Justice Center has analyzed data on ShotSpotter alerts in Chicago over a six-month period, from July 2019 through December 2019. The fundamental problem with the ShotSpotter technology is that it detects loud noises, gunshots among them. In a dense urban environment, this produces a high percentage of “false positives”: i.e., alerts that may or may not have been prompted by gunfire but lead the police to find no evidence of a gun crime or any other criminal activity. There are different ways of calculating false positives. Taking a conservative approach, the percentage of ShotSpotter alerts that resulted in no case report being filed was 85.35 percent.
The percentage of alerts that resulted in no case report being filed is not only evidence of unreliability but also a measure of waste. During the six-month period analyzed by the MacArthur Justice Center, there were 9,961 ShotSpotter alerts. Of these, 8,502 did not result in a case report. In other words, on 8,502 calls for service initiated by ShotSpotter police were called out to a specific location to investigate and found nothing worth reporting.
ShotSpotter is operative only in low-income Black and Hispanic neighborhoods and is coupled with software, also sold by ShotSpotter, that guides deployment decisions. The inevitable rejoinder will be: That’s where the crime is. Here, we encounter the circular logic of predictive policing by which supposedly scientific methods yield racist results, as overpolicing of communities of color drives an “evidence-based” dynamic that produces more overpolicing and attendant harms.
Those harms include the impact on targeted communities of the excess ShotSpotter-initiated calls for service that prove fruitless. Such interactions between police and community members are not only wasteful, they are also likely to be alienating after the fashion of blanket stop-and-frisk policies.
“Only people in the Black and Hispanic neighborhoods surveilled by ShotSpotter have to contend with the burden of thousands of unnecessary and potentially dangerous police deployments,” said Jonathan Manes, an attorney with the MacArthur Justice Center. “CPD’s use of ShotSpotter trades on a veneer of objectivity, but, like many high-tech strategies, the system ends up reinforcing racial disparities in policing.”
ShotSpotter dramatically increases the number of unnecessary police-community interactions and thereby increases the probability of bad outcomes that would not otherwise occur.
The large number of excess calls for service increases the probability of catastrophic encounters between police and community members. Again and again, incidents of police violence have arisen from relatively trivial occasions (e.g., a woman driving a car with a broken tail light, a man selling loose cigarettes, a child playing with a toy gun in a playground, et cetera). In view of the potential for any police encounter to derail, the first order of business is to reduce the number of unnecessary interactions. ShotSpotter does the opposite: It dramatically increases the number of such interactions and thereby increases the probability of bad outcomes that would not otherwise occur. This is all the more concerning in view of the aggressive manner in which officers, responding to what they believe may be gunfire, are likely to approach those they find at the location to which a ShotSpotter alert directs them.
Taken together, these findings are “shocking,” said Manes. “The ShotSpotter system in Chicago prompts thousands of deployments by police hunting for gunfire in vain. The system puts police on high alert, telling them that shots were just fired, but more than 85 percent of the time they don’t turn up evidence of any crime, let alone gun crime. These dead-end searches for gunfire happen nearly 50 times on an average day in Chicago. Each deployment is a powder keg situation for residents who just happen to be in the vicinity of a false alert.”
We don’t yet know what happened on the night police killed Adam Toledo. Perhaps officers arrived on the scene, found themselves under grave threat, and fired in self-defense. But it’s also possible that they rushed to the location of the alert and jumped out of their vehicle; that witnessing this, two individuals in a nearby alley ran away; that the officers pursued them and were carried to the fateful split second in which one of them shot the 13-year-old not by an immediate threat but by the blind momentum of the encounter.
Whether or not that happened on March 29, it’s all too easy to imagine it happening. In a society where there are more guns than people and in a state that permits concealed carry, it is arguably inevitable that there will be such outcomes, in view of the large number of excess police-community interactions prompted by “false positive” ShotSpotter alerts.
In response to a request for comment, spokesperson Sam Klepper stated, “ShotSpotter’s accuracy rate for detecting gunshots is 97% across the US for the last two years. This includes a false positive rate of 0.5%. Results are reviewed with agencies annually. We have more than 110 cities using ShotSpotter and extremely high customer satisfaction and renewal rates.”
This problematic technology is also extremely expensive. Chicago’s contract with ShotSpotter, which expires and is up for renewal in August, cost $30 million over three years. Then there is the cumulative cost of all the fruitless deployments in response to excess ShotSpotter calls for service. Above all, there is the incalculable cost of deaths at the hands of the police that could have been avoided.
In a Securities and Exchange Commission filing dated March 29, 2021 — the day Toledo was killed — ShotSpotter noted among risk factors for investors:
We may be adversely affected by ongoing social unrest, protests against racial inequality, protests against police brutality and movements such as “Defund the Police” or increases in such unrest that may occur in the future. These events may directly or indirectly affect police agency budgets and funding available to current and potential customers. Participants in these events may also attempt to create the perception that our solutions are contributing to the “problem,” which may adversely affect the Company, its business and results of operations, including its revenues, earnings and cash flows from operations.
ShotSpotter has reason to be concerned.