Predictive Policing Essay Sample

Policing requires data. Policing, the maintenance of law and order by a police force, often requires the use of data. In the text “Weapons of Math Destruction” by Cathy O’Neil policing algorithms such as PredPol that are used by under-funded police departments to track and prevent crime is critiqued due to the lack of data analysis. Specifically, it is the lack of interpretation of what O’Neil describes as victimless crimes. PredPol and many other software systems that track and predict crimes do not differentiate victimless crimes such as loitering, public drunkenness, and panhandling from felonies such as murders, rapes, and fraud crimes. This occurrence is known as “crime clusters”, O’Neil uses this phrase to describe inconsistencies  in the software systems used by police departments. Although algorithms may reduce arbitrary street policing, algorithms such as PredPol should be excluded from policing practices because it does not give an accurate picture of crime and is inherently discriminatory. 

American Police departments are notorious for poor policing methods. There are a variety of reasons, but one particular reason their police strategies are ineffective is that the data from the software system they use gives an inaccurate picture of crime. O’Neil agrees, she states “nuisance crimes are endemic to many impoverished neighborhoods. In some places, police call them antisocial behavior, or ASB. Unfortunately, including them in the model threatens to skew the analysis, page 76.” This statement illustrates how nuisance crimes are included in the data that police used to determine which neighborhoods to police, skewing the data, painting a picture of crime that doesn’t exist. This might not seem like an issue large enough to get rid of predictive crime software, but this inaccurate painting of crime is not harmless; it often causes a continuous cycle of policing. O’Neil states, “This creates a pernicious feedback loop. The policing itself spawns new data, which justifies more policing”, page 76. This statement explains how the lack of data analysis leads to over-policing, more specifically feedback loops. A feedback loop is the part of a system in which some portion or all of the system’s output is used as input for future operations. Predictive policing software misses the mark when it comes to analyzing data, causing feedback loops and inaccurate conclusions overall. 

Although many academics agree that predictive policing deserves the criticism it receives, others don’t, some argue that it works to deter crime. A quote that they use is “When police in the British city of Kent tried out PredPol, in 2013, they incorporated nuisance crime data into their model. It seemed to work. They found that the PredPol squares were ten times as efficient as random patrolling”, page 77. It may be true that predictive policing is more efficient than random patrolling but critics of predictive policing software are not advocating for officers to troll the streets with no agenda, they are advocating for police departments to stop grouping misdemeanors and felonies in the same categories causing data to be biased. Predictive policing supporters support their claim that PredPole and similar software should be used by using the quote “if cops spend more time in high-risk zones, foiling burglars and car thieves, there’s good reason to believe that the community benefits page 76.” It’s easy to think this way but facts disagree, police officers often spend their time arresting people committing petty crimes not felons when they use policing software like PredPole. The last quote supporters use to justify their belief is “a number of the crime prediction models are more sophisticated because they predict progressions that could lead to waves of crime” Page 75. At first glance this argument makes sense but it is inconsistent with research, Policing software can’t interpret what data means making it a poor tool to predict crime. So it’s not as sophisticated as it’s marketed as. 

More Americans are aware of the systemic racism and classism that plagues the nation’s police departments. Predictive software is partially to blame. The software targets specific neighborhoods which are disproportionately populated with marginalized people. Weapons of Math destruction states “And in most jurisdictions, sadly, such a crime map would track poverty“ page 78. This quote describes how the majority of predictive software ends up targeting impoverished neighborhoods and not areas riddled with violent crime. This is significant when it comes to discussions about justice and equality amongst economic classes. W.M.D also states “Most of them come from impoverished neighborhoods, and most are black or Hispanic” page 76. This statement highlights the blatant discrimination predictive software is responsible for. Minorities are disproportionately targeted when police software is used causing more data to enter the system causing the public to believe that minorities commit the most crimes despite making up a small population of the country. Page 82 of W.M.D states “So fairness isn’t calculated into WMDs. And the result is massive, industrial production of unfairness.” This quote explicitly says that this system is unfair. Unfairness is a part of life but this particular type of unfairness leads to a web of data that are based on unfair conclusions causing issues bigger than the original unfair collection of data. Page 76 of W.M.D states “And our prisons fill up with hundreds of thousands of people found guilty of victimless crimes.” This quote brings the flawed criminal justice system to light and how predictive policing contributes. The majority of people in the criminal justice system are black and brown, predictive policing is just another way to imprison black and brown people for minuscule crimes that involve no victims.

Overall supporters of predictive software for policing make a point in saying it could reduce arbitrary policing however, policing software like PredPole should be excluded because it does not illustrate an accurate picture of crime and is inherently discriminatory. Predictive software does not differentiate the different types of crime, this leads to police believing an area is overwhelmed with violence. Predictive software also is partially responsible for overcrowding prisons with black and brown people since these neighborhoods are populated with minorities and poor people. For these reasons, Police departments should take PredPole and similar software out of their jurisdictions.

Works Cited

O’Neil, C. H. Weapons of Math Destruction. Penguin Books, 2017.