Crime rates 2019: Inside real-life ‘Minority Report’ predictive policing slashing crime


CRIME RATES could take a drastic cut thanks to an innovative form of predictive policing, an expert has revealed.

Pre-crime is the disturbing concept set out in the dystopian Philip K Dick novel Minority Report. But predictive policing is already a reality and preventing crimes before they actually occur. And the latest cutting-edge crime-fighting tool for police forces around the world is not weapons or radio – it is data.

Experts have grappled with the ramifications of “pre-crime” ever since Minority Report, which was later turned into a Tom Cruise film.

Pre-crime is the idea police can potentially predict when an offence will occur to take measures to prevent it.

PredPol, described as the market leader in predictive policing, works with 50 police departments, representing roughly 20,000 total patrol officers across the US, and has even been used in the county of Kent.

PredPol CEO Brian MacDonald told how a data-driven approach is slashing crime rates.

He said: “We find roughly 80 percent of police agencies don’t provide any kind of explicit guidance on officers on patrol.”

“There is not a whole lot of guidance and it’s not scientific.

“The biggest problem is officers’ potential biases.

“The best way to predict what crime is going to occur is where is to look at a historical pattern of work crimes have occurred previously and then understand the kind of story those crimes tell you.

“We’re trying to take the officer’s judgment, and the potential for bias, out of the equation of control operations.”

Although he admits it is hard to qualify Predpoi’s efficiency in cutting crime, he believes using the company “generally double digit decrease in targeted crimes.”

Mr MacDonald said: “We have one agency in California that saw about 50 percent reduction in auto theft over a five year period, which is pretty significant.

“We have one in New Jersey that saw a 60 percent in robberies in one year – again, very significant.

“My chief scientist is always warning me you can’t prove cause causation in social sciences generally, but you can prove a pretty strong correlation.

“So we found when they roll out in one part of town but not the other, the crime rates go down where they rolled it out, but it hasn’t gone down the parts where they haven’t used PredPol.

“What we do find is an agency that uses PredPol assiduously will see significant decline in the crime rate.”

He is at pains to stress how crime reduction is not due to an increase in arrests, but because police are interfering with more opportunities to commit crime.

He said: “The goal is really not to catch bad guys – the goal is to prevent the crimes.

“You’re preventing the victimisation and you’re preventing somebody getting arrested and going through the criminal justice system, so it’s a kind of win-win.”

The company claims to incorporate up to five years of historical data which is then parsed through a machine learning algorithm.

In a bid to reassure people police will soon identify people before they break the law, he added: “We never use any personally identifiable information.

“It’s all about location, so the only data we use when we do our analyses is crime type, crime location and crime date and time, just those three data points.

“So we never talk about who is likely to commit that crime or who’s likely to commit the next time.

“If you want to impact crime go to these locations during a patrol and and you’ll help deter those crimes from happening by your presence.”

We’re trying to take the officer’s judgment, and the potential for bias, out of the equation of control operations


About Author

Leave A Reply