Predictive Policing Used for Crime Prevention


Some United States criminal justice organizations are experimenting with technology to identify and prevent illegal activity. Some law enforcement agencies use predictive policing software with the goal of forecasting the people and places most likely to contribute to future crimes. This software uses crime data from local and national sources.

Depending on the desired forecast, the agencies can analyze information using various processes such as predictive intelligence and reoffender forecasting. Around the country, law enforcement agencies are expressing interest in exploring the possibility of substantial performance improvements through the use of predictive police technology.

Policing in the 21st Century

According to research by the RAND Corporation’s Safety and Justice Program, big data analysis technology tools provide law enforcement agencies with unprecedented forecasting opportunities. The underlying framework originates directly from the business use of technology to forecast consumer and market trends. Law enforcement use of the technology, however, would be for forecasting crime patterns, improving tactical efficiency and allocating resources to high-risk targets.

Creating a Comprehensive Picture

Predictive policing analyses can take several forms. According to researchers in RAND Corporation’s Safety and Justice Program, the most common are:

- High-risk location and time forecasting
Using risk terrain analysis, law enforcement agencies can identify high crime-risk places, geographic variables and times.

- Potential offender risk
Regression and classification modeling uses previously identified risk factors to predict offending risk.

- Suspect identity matching
Computer-assisted queries help identify individuals associated with other offenses resembling active cases, or compare evidence generated through various sources such as GPS tracking records or traffic camera images. Tools like geographic profiling tools and crime linking analysis can be used to seek the identities of perpetrators.

- Victimization risk forecasting
Crime patterns can be used to identify hot spots where individuals, businesses, and communities most at risk for crime overall and of particular types.

Examples of Predictive Policing Deployments

In Shreveport, Louisiana, law enforcement officials are experimenting with the Predictive Intelligence-Led Operational Targeting (PILOT) program. Using reports involving lesser crimes gathered from the immediate and surrounding coverage areas, the agency forecasts the increase in short-term illegal activity and then intervenes by increasing law enforcement presence in the predicted high-risk areas.

Using this method, the officers monitor the high-risk areas for potential criminal activity. While building intelligence gathering connections and rapport, the officers network with community members and explain the PILOT program. Using the resulting information, Shreveport law enforcement researchers develop crime maps and situational awareness dossiers. Agency commanders then are able to target high-risk areas for arrests and information gathering, with the goal of reduced criminal activity rather than increased arrests.

The New York Times reports that the Kansas City Police Department uses predictive policing software to monitor and analyze criminal organizations, social media activity and drug consumption. Through an initiative called the Kansas City No Violence Alliance (KC NoVA), officers use the resulting information to directly confront and challenge potential offenders during mandatory group call-ins. The agency uses the meetings to inform former offenders that the department is monitoring their activities, hopefully discouraging the individuals from repeating criminal acts.

KC NoVa also actively assists potential reoffenders by providing important resources, such as connections to potential jobs and supportive housing. Since implementing the initiative, the department has successfully reduced homicides by 20 percent. The Kansas City Police Department has also successfully identified almost 60 criminal enterprises with nearly 1,000 members. Among the most encouraging outcomes; some call-in participants make positive contributions to the same communities in which they previously perpetrated their crimes.

Continuing Growth of Predictive Policing Technology

Analytic tools that help law enforcement agencies understand crime patterns hold promise in helping departments be more effective and efficient with their limited resources. Targeting the highest risk people, places, and times provides greater opportunities to reduce crime, compared to more reactive approaches which respond to criminal events that have already occurred.

To further these efforts, researchers and law enforcement agencies have developed partnerships aimed at making crime data more available and understandable to the public. One such partnership formed between Portland Police Bureau and Portland State University (PSU). The Bureau is working with the University’s Criminal Justice Policy Research Institute, within PSU’s Criminology & Criminal Justice Department, to promote the public awareness and increase strategic use of crime and public safety data to target specific crime problems in the community.

It is important for people to understand how these analysis techniques benefit their communities as well as to have a realistic understanding of crime in their area. As general data processing and analysis capabilities improve, it is likely that the use of predictive policing approaches will continue to expand around the country.

Learn More:

The criminal justice field needs innovative leaders. Our online criminology and criminal justice degree at Portland State University can help you become one. Our rigorous curriculum and award-winning faculty give you the opportunity to examine criminal behavior in-depth, build advanced crime analysis skills, and develop strong leadership characteristics.