Person of Interest: Chicago’s Predictive Policing, How Does it Work?
Many 1984 predictions have somewhat come true in this day and age, and a recent uncanny reminder of the Orwellian world is predictive policing. Akin to the CBS drama’s Person of Interest’s main characters, police departments in the United States have started using predictive policing software to help reduce crime rates in urban areas, including Chicago, Memphis, and Atlanta. Cities use different systems, but they are similar in terms of the variables that affect the outcomes: individuals’ arrest history, acquaintances and their arrest history, behavior while on parole etc. How is this working?
Despite employing predictive policing techniques, murder rates have not gone down in the city of Chicago over the past 4–5 years. The CPD’s system aims to target people that are potential criminals or potential victims, warn them based on their findings and offer measures that may help avoiding the predicted murder/shooting. Grosso modo, they look at individuals’ records and note who they’ve been arrested *with* in the past, whether they’d ever been present on a crime scene, on the victim or criminal’s side and other factors along those lines.
Theoretically, predictive policing sounds great, especially when paired with measures that give people in crime-ridden communities the opportunity to defy the odds and succeed per society’s standards. In Chicago, once someone appears on the “hot list,” A police officer visits their household (based on address on file), *sometimes* accompanied by a community leader and or social worker, whose purpose may be bridging the gap that exists between police and community. Individuals are warned about being on this “hot list” as either potential victims or criminals, and are offered entry into programs that would help them acquire a high school diploma and/or further their education. The first visit is usually followed by a phone call and in some cases a second visit, and then the case is closed. In the case that the person of interest is a potential criminal, the officer insists on the fact that the police department takes the threat seriously and that they will get prosecuted and sent to jail if they commit the crime. Sounds more like threats of arrest on no grounds to me, and likely to anybody from a community that doesn’t have good relations with the police.
Based on the description of their predictive policing practice, the CPD has failed over the years with its techniques on two fronts: cultural competency and thoroughness when it comes to contact with the person of interest. Taking into consideration the fact that people that have a history of arrest, drug abuse etc may not have the friendliest relationship with the police, it is up to the police department to create an environment where people from all walks of life feel comfortable having these discussions. Although the chief of the CPD notes that the software they use does not take location into consideration, there are specific pockets of crime in the city of Chicago, and the police department could put efforts and resources into fostering better community relationships in collaboration with the mayor’s office (stop closing schools, by the way.) In areas such as the south side of the city where most individuals arrested and most victims are people of color, having activities that involve the youth, the leaders of the local community (read clergy, educators etc) and the police would foster tighter bonds and eventually, even if only in the long term, erode the negative responses and lack of interest in programs the police officers may receive from the persons of interest.
When describing the steps taken by police officers and the social worker in order to inform the persons of interest, it appeared as though informing a family member in lieu of the person in question sufficed as a step towards reducing the risk of crime occurring. What the CPD is failing to realize here is that in many instances, individuals involved in criminal/gang activities may not have the greatest relationship with their family members. This puts into question the effectiveness of this side of the practice. It seems as though the general attitude from the predictive policing side of things is “we warned them and gave them a list of options, it’s not our problem now! *Kanye shrug*”
The bottom line is, it is wasteful to spend lots of money on finding out who is likely to be killed and who will likely commit a crime, yet not invest enough time, efforts and resources towards the actual prevention methods. This makes people wonder whether the CPD is more interested in being prepared to arrest even more people, than actually saving lives and improving the conditions in the community.
Listen to NPR’s “On Point” Episode on predictive policing
Read this Pro-Publica article on Machine Bias: how predictive policing software is intrinsically biased against people of color.