Predictive Policing: Can Cops Anticipate Crime Through Data?

Predictive policing is the result of crime analytics and the tools used to collect data by law enforcement.
Photo Courtesy of Shot Spotter.

Imagine a police force armed with advanced data analytics, that can anticipate criminal activity and work to prevent it before it even happens.

Actually, it doesn’t have to be imagined. Author Philip K. Dick already did so in a short story that was the source material for the movie, “Minority Report,” which centered on predictive policing.

Of course, in that story police used information from people who could see the future. That sort of fantasy no longer is needed. Data analytics can provide the same service.

Predictive policing involves leveraging information from large data sets to identify patterns in criminal activity. Police use these patterns to allocate resources that will prevent or mitigate the impact of crime. They can even identify those most likely to become criminals and victims.

Several municipalities are now moving toward adoption of crime analytics tools. One high profile example in recent years is the adoption of gunfire detection technology, which can help police create maps of criminal activity involving the use of firearms, as well increase reaction time to an incident.

An example of a gunfire detection technology incident map. Photo courtesy of Shot Spotter Inc.

Predictive policing could prove to be the story of law enforcement in the 21st century. It’s already in use in dozens of police departments and district attorney offices, from Los Angeles and Miami to New York City.

Implementing Data

Crime prevention has long been a focus for law enforcement. Using primarily “street intelligence” – i.e., information given by informants – law enforcement attempts to act before a crime occurs. Data analytics has added a new dimension to that approach.

Street intelligence remains important. But now, law enforcement can combine that information from insights gained through data analysis like never before. Crime analytics can include any or all of the following:

  •          Criminal records
  •          Social media activity
  •          Known associates and friends
  •          Past drug use
  •          “Hot spots” where many crimes occur
  •          Determining what crimes occur more frequently in certain locations based on weather conditions

Taking this data and combining it with long-practiced strategies such as closely monitoring those recently paroled can increase police presence in the areas where it is needed.

Police (or consultants hired by law enforcement) crunch the numbers. They then produce predictive maps that show when and where a crime is expected to occur. Police increase patrols in that area. In some cases, they may take special care to watch specific individuals.

The information also can be used to warn people in certain areas that they are more likely to become victims of crime, even going so far as to offer them social services and other help.

The Potential and the Criticism

Many police agencies say predictive policing works. For example, in Pittsburgh police are using data from Carnegie Mellon University to allocate patrol resources in areas where they feel crime is most likely to happen, including the Homewood section of the city. It is an area where homicides frequently occur.

In Chicago, police have created a list of 400 people far more likely to be involved in crimes than others. In Kansas City, homicide rates have dropped 20% after police began using predictive analytics.

Those involved with crime analytics say that criminals tend to operate in distinctive patterns in terms of what crimes they commit and when and where they commit them. That offers police a chance to better prevent crimes.

However, others are not so sure. In an analysis by the RAND Corporation, the advantage of predictive policing over other long-held practices was found to be slight. Human rights groups also have argued that predictive policing can lead to people coming under scrutiny by the police because of guilt by association.

Ezekiel Edwards, with the American Civil Liberties Union, told the New York Times that suspicion is created around people because they “live in a certain neighborhood or hang out with certain people.”

Expect police to continue using a mix of crime analytics, community policing and old-fashioned techniques to determine where crime will occur – and for there to be continued debate over whether marking someone as a potential criminal is fair in a democratic society.


Please enter your comment!
Please enter your name here