Is Predictive Policing Like Minority Report?

predictive policing

By Clarissa Manning

Imagine the ability to foresee every crime before it happens and stop the expected perpetrator before any heinous act is committed. If this scenario sounds familiar, you’re probably thinking of the 2002 film Minority Report. In the non-cinematic world, the use of mathematical models based on risk factors and past criminal behavior to predict future criminal activity is commonly referred to as predictive policing. This relatively new technology often draws comparisons to the sinister “Precogs” found in that film. This characterization, however, could not be further from the truth.

Crime mapping, the precursor to modern predictive policing, dates to the 1800’s. French lawyer Andre-Michel Guerry and Venetian geographer Adriano Balbi created maps depicting the density of personal and property crimes in the departments (sub-regions) of France. This primitive technique sufficed for over a century, until the invention of computers. Computers allowed police departments across the world to utilize highly sophisticated crime-mapping technology, commonly referred to as hot-spot analysis, to assist in determining where limited policing resources can best be allocated.

Today, crime analysts spend much time analyzing crime trends via mapping techniques to determine where crimes are likely to occur. Over the past decade, though, new mathematical algorithms designed to provide police with more nuanced predictions of criminal activity have been tested in the field alongside the traditional hot-spot maps. The results are promising.

The Los Angeles Police Department currently uses software similar to that used by seismologists measuring the risk of seismic activity. Using only the type, place, and time of previous crimes, this software develops predictions of criminal activity. This technology is intended to produce more accurate and predictive hot-spot maps, and ultimately reduce crime. A recent study found that it does just that; police patrols using predictive forecasts led to an average 7.4% reduction in crime volume as a function of patrol time, whereas patrols based upon analyst predictions showed no significant effect at all.

The implications of this software are not without concerns, however. As police departments increasingly adopt predictive policing technologies, courts must weigh the reasonable suspicion claims derived from this technology against the  4th Amendment prohibition of unreasonable search and seizure. The courts have found that a suspiciously-behaving individual who avoids police in a high-crime area meets the criteria for reasonable suspicion.

But this is an extremely narrow and specific circumstance. You might consider an alternative situation in which an individual carrying a tool bag in an area predicted to experience crime in the near future is stopped. Does this meet the reasonable suspicion requirement? As predictive policing technology accounts for the environmental characteristics that allow crime to occur in an area, so should the courts. If the zone of predicted criminal activity in which this individual had been stopped was previously a poorly lit alleyway that was subsequently outfitted with several streets lights and security cameras, does a claim of reasonable suspicion still exist? What obligation do the police have to report to the courts a change in the environment that may negate the prediction of criminal activity? These questions remain largely unanswered.

Legal issues aside, the use of crime mapping is a vital tool for police departments, and the use of predictive policing technology continues to play a role in crime reduction. The importance of this technology, however, should not negate the responsibility of officers and the courts to protect the fundamental rights of individuals.

104 queries in 0.839 seconds