Santa Cruz police spend an average of 40 minutes of every hour on emergency calls, leaving them with 20 minutes to patrol the streets. For the past few weeks, some have been using computer mapping systems to help them determine where to best spend the little time they have left. The program’s success has drawn notice from the LAPD, whose officers hope the technology will save time and money in scheduling patrols and preventing crime.
In law enforcement circles, the catchphrase for the futuristic tool is “predictive policing.“ Police departments have been providing years of historic crime data to mathematicians, who've created algorithms to analyze and determine crime patterns. The results are predictions of where and when similar crimes are likely to occur.
Zach Friend, a crime analyst for the Santa Cruz Police Department, says the crime-fighting system is modeled on methods for predicting earthquake aftershocks. The tool comes from Santa Clara University Professor George Mohler who believes crimes follow similar patterns. Friend, who helped to launch the program in Santa Cruz, says the system works because crimes tend to occur in time and place-based patterns. Santa Cruz officials became interested in the program after the success of a similar pilot by the LAPD.
“You have a crime and there will be after-crimes that occur after that,” said Friend. The technology, he says, has helped Santa Cruz prevent crimes before they happen. Thus far his department has focused on burglaries and vehicle theft.
In L.A., LAPD Captain Sean Malinowski says he'd like to push the envelope further; and next year use the technology to predict violent crimes. Each morning officers using the program enter crime reports into the system, which is already packed with eight years worth of data. The program then predicts 10 potential crime hot spots.
Malinowski says the technology represents a vast improvement to what the department currently uses.
“The instruments we are using seem blunt now, in terms of the kind of specificity we can get with data analysis,” he says. Malinowski says he believes the computer model helps to remove biases.
Police in Santa Cruz say they have seen real results from the program’s predictions, which often result in arrests or crime prevention. However, Marjorie Cohn, professor of law at Thomas Jefferson School of Law, says she is concerned about the consequences of predictive policing.
“It's very worrisome to me. By directing police patrols towards these locations, [police] can harass, detain or arrest people more efficiently,” she says. “In certain neighborhoods people run from police even if they've done nothing wrong, and that's because of past harassments."
Malinowski says they are addressing such concerns. His plan includes consultations with civil rights groups to ensure communities will not be targeted unfairly. He also says L.A.’s predictive crime pilot program will be highly scientific, and include a controlled environment and test site to measure the program's success.
Stanford Law professor Robert Weisberg says there's still work to be done, “We are we are a zillion miles way from having enough data and analysis to evaluate just how well it works.”
Even so, the results have already convinced Friend. He is adamant predictive policing is the future, or at least part of it.
“It will be [the future] for a number of reasons. [Police] staffing is down, yet we still need to patrol the same amount of ground. This costs us nothing. And there is no huge downside,” he says.
Zach Friend, Crime Analyst, Santa Cruz Police Department
Sean Malinowski, Police Captain, Commanding Officer, Foothill Patrol Division, Los Angeles Police Department; Principal Investigator on the National Institute of Justice funded “Los Angeles Predictive Policing Planning Project”
Marjorie Cohn, Professor of Law at Thomas Jefferson School of law; Co-Author, "Cameras in the Courtroom: Television & the Pursuit of Justice"
George Tita, Criminologist, University of California at Irvine