Estimated read time: 4-5 minutes
CHICAGO (AP) — Chicago police have ended a program that sought to predict people most likely to be victims or perpetrators of gun crime, according to a watchdog report released Thursday that warned the department of serious flaws in the approach.
The police department implemented the method in 2012, using a formula developed by researchers at the Illinois Institute of Technology and backed by $3.8 million in federal grants. Department spokesman Luis Agostini confirmed Thursday that the department decided to end the program after researchers at the nonprofit RAND Corporation reviewed it last year.
The formula created a list of people deemed most likely to commit or be victimized by a shooting. It relied on several factors including age during an individual's latest arrest, the number of times someone was the victim of a shooting, battery or assault and their total number of arrests for unlawful weapons.
Early versions of the formula also used gang affiliation and narcotics arrests to calculate risk scores; the report says those criteria were later removed.
The higher your score, the more likely you were considered to become a victim or perpetrator of gun violence under the “Crime and Victimization Risk Model," previously known as the “Strategic Subject List."
Police departments around the country have been criticized for similar programs aiming to predict where crime will occur and who is likely to commit it, particularly by civil rights advocates who fear the “predictive” policing strategies will prompt unwarranted police attention on minority communities.
In Chicago, the program became an often cited part of a police strategy to combat violence, particularly after a spike in violent crime in 2016 including 773 homicides.
The scores were used by the department's “custom notification” program, which sent police and outreach workers to meet with people deemed to be at risk of becoming a victim or perpetrator of gun violence.
But the scores also appeared in arrest records and were used to identify people for targeted busts. The Chicago Sun-Times reported in 2017 that the list included people who were never charged with a violent crime or illegal gun possession after a lengthy legal battle with the department to obtain a version of the list.
Inspector General Joseph Ferguson’s office said the police department provided notice in August that it planned to stop using the program when federal grant money ran out a month later. According to the report, 313,513 individuals had been scored using the model by March.
Ferguson's office's found the program was too dependent on arrest records, making some scores unreliable. For instance, the report said someone arrested for a non-violent misdemeanor would be scored while someone wounded in a shooting but never arrested would not be.
Relying on arrest records also “effectively punished individuals for criminal acts for which they had not been convicted," the report said.
Ferguson's office found the police department failed to train officers and commanders about the meaning of the scores and failed to update the totals between August 2016 and January 2019.
"While these particular risk models have now been decommissioned, there are critically important and widely applicable lessons to learn here about the importance of careful data handling and thoughtful, purpose-driven policies and training, in ongoing and future efforts to harness data that predicts and ultimately prevents violent crime,” Ferguson said in a statement.
The researchers, who noted they did not have detailed information about the formula, concluded that the program may be capable of identifying people at higher risk of becoming victims of gun violence. But they noted that the scores quickly get outdated and are time-consuming to keep updated, ill-suited for a police department seeking to use the information to prevent violent crime.
"We will continue investing in technology that supports our collective mission," Agostini said.
He said that includes the department's network of tech-focused facilities, known as Strategic Decision Support Centers. The centers are staffed by police and data analysts, who use surveillance video, gunshot detection systems and other information to help guide patrol officers on the streets and develop strategies to prevent gun violence.
The American Civil Liberties Union of Illinois said the inspector general's findings demand more disclosure about technology still in use by the department.
“Transparency is critical to the ongoing effort to create faith in policing in Chicago,” said Karen Sheley, director of the organization's Police Practices Project. "Unregulated, sophisticated, powerful databases involving hundreds of thousands of names only harm that trust.”