The Beginning of PreCrime?

What is Predictive Policing?

Predictive policing is an attempt to make police more proactive. Often, police react to crime by responding to 911 calls or arresting people who they see break the law. However, new technology allows police to prevent crime before it starts. Large amounts of data about past crimes are fed into computer programs that determine where and when (and sometimes with whom) the risk of crime increases.

Since TIME magazine called predictive policing one of the “50 Best Inventions of 2011,” more towns and cities are embracing this new technology.

Geographical Predictions

Predictive policing based on mapping has been very successful in combatting crimes like auto theft that occur frequently (and so create a lot of data) and follow certain patterns. Officers generally receive a map each day with highlighted areas to patrol when they are not responding to calls. In some places, patrolling the high-risk mapped areas is not mandatory. However, it is encouraged by departments because it gives officers more information. 

Many U.S. cities contract with PredPol, a private company that takes crime data and uses a patented process to make predictions. In Santa Cruz, California, the police use the tool to map potential gang violence, batteries, aggravated assaults, drug crimes, and bike thefts. In Los Angeles, the police chose specific areas to experiment with Pred Pol and waited for the results to come in. In the Foothill area, the experiment was successful and property crimes fell by 12%. Now, the LAPD has adopted it in 14 out of 21 of its divisions. It is estimated that the prevention provided by this technology could save citizens $9 million dollars a year.  One critique is that the system does not work as well with crimes that don’t “have enough data points” or that involve “crimes of passion.” For these reasons, homicides are more difficult to predict on a map.

Individualized Predictions

Recent advances in predictive policing mean that it is also being used to target violent crime. Since violent crime typically involves individuals instead of property, it is no surprise that police departments are now tracking specific people as part of their “prediction” strategy. These predictions are often based in theories about criminal psychology. For example, the Chicago Police have developed a list of 400 people most likely, at the given time, to commit a violent crime. To do so, they use social media data, crime information, suspicious person reports, and 911-calls. However, this so-called “heat list” has led to criticisms by civil liberties groups like the ACLU. The risk is that predictive policing increases police scrutiny on individuals regardless of whether or not they have committed a crime.

Other jurisdictions are also embracing a more individualized risk model. In Maryland, for example, social service workers are teaming up with statisticians to predict which families will be most likely to seriously harm their children. Meanwhile, in Kansas City, the authorities have identified almost 1,000 people belonging to criminal groups and are targeting them specifically with interventions intended to curb violent crime.

What are the Criticisms of Predictive Policing?

While map-based predictive policing is certainly effective at deterring some types of crime, civil liberties activists are concerned that it may target residents of certain neighborhoods unfairly. If utilized in the wrong way, this type of policing may create a vicious cycle.

For example, racial profiling or increased patrolling of certain neighborhoods may lead to black or Hispanic individuals’ arrest for drug crimes at a higher rate than their white counterparts (who commit a similar number of those crimes). Data from these arrests may create a false impression that there is more of a “drug problem” in predominantly black or Hispanic neighborhoods, leading to increased patrols, more scrutiny, and more arrests. A similar problem may involve patrolling in “rich” versus “poor” neighborhoods.

Targeting individuals may be even more problematic. It sets up a new and invasive type of surveillance, not over neighborhoods in general, but over people in particular. Now, license plates are scanned, devices can track mobile phone signals, and surveillance cameras and tablets with facial-recognition technology are now available to police. This makes a very detailed level of tracking attainable, and moves us closer to a “Minority Report” world.

What is the Future of Predictive Policing?

Predictive policing holds great promise as a deterrent to certain types of crime. However, the use of predictive policing must continue to be scrutinized carefully so that it does not interfere with important civil rights. Many cities are now experimenting with this technology, so society will eventually know its effects, good and bad.