Photo courtesy of Wikimedia Commons, Author, Camelia.boban
The image of a policeman seated in front of a crystal ball or a magical mirror does not seem logical, even to the most creative minds. Nevertheless, the science fiction movie “Minority Report” shows a police department that arrests people who are expected to commit crimes in the future. More recently, the CBS television series “Person of Interest” shows a computerized vigilance system that secretly monitors everything and uses algorithms to predict which people will be involved in a violent crime.
Although there’s a vast gap between fiction and reality, that is the general concept of what predictive policing means. It is a way of using advanced technology and data analysis to adopt preventive measures, to “get ahead” of the crime. Thanks to the new technologies, predictive policing may be able to reduce crime by carrying out patrols and operations based on reports generated by algorithms that analyze data on past crimes.
These methods are very different from the more traditional work of police forces. Even in those countries with the most innovative police forces, it’s clear that most of the work of law enforcement is reactive: they answer calls and complaints, they send officers to locations and they initiate investigations with the goal of carrying out needed arrests. Unfortunately, these reactive methods do not generate higher levels of security for citizens or higher levels of trust in the police and do not prevent new crimes – even when the criminal is arrested and jailed. Today, more than ever, law enforcement requires a proactive approach, and that’s where predictive policing is successful.
In a nutshell, the predictive policing model means using the information we already have to try to anticipate future crimes. When we speak with veteran police officers, we can see that the logic of predictive policing is not new to them. They even say that with some limitations, they themselves used similar methods for years. Those practices make sense because decades of anecdotal evidence show that when there’s a crime in one part of the city, the odds increase that neighbors will be victimized as well. Veteran police officers also know that criminals prefer to operate in certain parts of the city and have certain patterns of activity for certain hours and days of the week.
What is new today is the volume of data that can be processed with new technology and the speed of the analysis of that data. It’s no longer a police officer’s intuition, but scientific evidence. Today’s law enforcement can develop an algorithm (a mathematical process) that calculates the likely locations of future crimes. The algorithm can process data on the anthropological and criminal behavior of different groups of people. In short, predictive policing uses complex mathematics to estimate the likely hotspots for future crimes.
It is important to point out that analytical methods used in predictive policing do not identify or seek to identify individual persons. The emphasis is on obtaining relevant data on the hours, places and types of crimes that may take place. With this data, police officers may carry out preventive operations, reassign human resources and evaluate the best possible responses by authorities. These models also help to assign the economic resources of law enforcement in a more efficient manner. In the long run, they also help to develop education and training plans and the establishment of broader public policies.
One of the more detailed studies of the concept and ways of implementing predictive policing systems was carried out by the RAND Corporation. A series of studies in the United States also documented the positive impact of predictive policing on crime reduction in various U.S. cities.
Predictive policing is no easy matter, however, and by itself cannot replace all the tactics and strategies used by effective police forces. The accuracy of the predictive policing programs depends on the accuracy of the information entered into the systems. The programs, like the majority of social-related programs, also may lack proper elements and scientific validity.
Follow us on Twitter: @IDB_CarDevTrend