Jacquelyn Damon spends her time with a powerful crime-fighting tool at the Santa Clara County Sheriff’s Office. It doesn’t fire projectiles of any sort or turn concrete transparent from helicopter heights. The former high-tech product manager, now an analyst, works to beat the bad guys with an Excel spreadsheet.
The process, known as predictive policing, uses mathematical formulas to prevent burglaries and thefts before they occur.
At a time when cities everywhere are struggling under the pressure of budget shortfalls and being forced to cut essential services, the promise of a technology with the ability to improve the efficiency of public safety without threatening jobs is so attractive it almost borders on intoxicating.
It’s a sexy concept. Last summer, the Santa Cruz Police Department embarked on its own six-month experiment to see if it could reduce crime by proactively deploying officers to patrol areas deemed by a statistical algorithm to be the most likely spots for a crime to occur. It was the first department in the country to put this theory into practice, and the move attracted national attention.
The story landed in The New York Times and on the cover of Popular Science. Time named it one of the 50 best innovations of 2011—all before the program had any real results to provide.
Could technology improve public safety and cost nothing while maximizing resources left intact after severe budget cuts? The methodology can’t be accused of unfairly profiling by race or socioeconomic standing, because the only things the computer cares about are the crime’s location, time and type.
Oddly enough, Santa Cruz‘s more heralded program didn’t come close to matching the dent in crime percentagewise that Damon’s predictive policing model—funded by the U.S. Department of Justice—showed in its first full year.
“The most common time [vehicle and residential] crimes were occurring were Tuesdays and Thursdays between 5pm and 8pm,” says Damon, who works out of a sheriff’s office substation in Cupertino. “We put together hot spots and victim profiles to give officers an idea what to look for. In May of 2010, we started seeing a significant decrease. It was pretty immediate once we got our patrol units in the right place at the right time.”
As a result, from 2010 to 2011, property crimes in the West Valley patrol area for the sheriff’s office—Cupertino, Saratoga, Los Altos and unincorporated zones that include parts of Los Gatos—dropped 23 percent, according to Damon.
Meanwhile, during the first half of 2011, Zach Friend, a spokesman for the Santa Cruz Police Department, says that after using its predictive policing algorithm, the department reported a drop in property crimes ranging somewhere between 4 and 11 percent.
Breaking down Santa Clara County’s stats, Lt. Don Morrissey says West Valley cities that contract with the sheriff’s office for law enforcement services saw declines in property crimes: 16 percent in Los Altos, 18 percent in Saratoga and 34 percent in Cupertino. Morrissey largely credits Damon’s spreadsheets, which are updated and disseminated to patrol sergeants once a week.
“I can’t say enough about the work she does,” he says. “Having that resource here, and allowing us to sit back and take a closer look at the time these crimes are happening and disseminate that amongst the deputies, whether they’re working here all that time or just for that day, everybody is working off of the same information.”
“It’s funny, I like to say I was always a product manager,” says Damon, who handled product launches for companies that contracted with Cisco before joining the sheriff’s office as a communications dispatcher in 2004. “It’s just my product is crime now instead of technology.”
The model being tested in Santa Cruz is an outgrowth of a research project by George Mohler, now an assistant math professor at Santa Clara University. “Self-Exciting Point Processes Explain Spatial-Temporal Patterns in Crime” was the name of his paper, published last March in the Journal of the American Statistical Association. Mohler conducted the project at UCLA, where he was working as a postdoctoral researcher.
The Mathematical and Simulation Modeling of Crime research project at UCLA is funded by federal money from the National Science Foundation and headed by anthropologist Jeff Brantingham. Last spring Mohler was working on the project, which uses math to model crime statistics in space and time. Time-lapse versions of these maps show crime moving across a map like weather patterns across a blue screen.
“We all believe that we have lots of deterministic control over what we do and where we do it and how we do it and why,” Brantingham says. “But at the same time there are also a lot of aspects of human behavior that are actually very general and can be described very effectively in simple ways.”
That could explain why Damon’s model is having such success. “You can do it with an algorithm or you can use pen and paper and be just as successful,” says Damon, whose position, along with two other crime analysts for the county, was federally funded to run through June 2013 at a cost of almost $800,000. “It’s probably not as efficient as what Santa Cruz is doing, but what I’ve been doing is temporal and spatial patterns. It’s not so much stopping people. Mostly, it’s just being a presence in the area where we get these types of crimes.”
But when a string of auto theft burglaries hit the Foothill area last summer, Damon says she prepared a report on July 23 that documented times and places for patrol units to focus more efforts. Within a week, a patrol check of the neighborhood put the sheriff’s office in contact with a parolee strolling the block late one night. Days later, a search warrant was conducted at the parolee’s house and items from at least four burglarized cars were located.
When observed from a distance, Brantingham says, there is not much difference between the ways people behave and the ways fluids or gas molecules behave. With enough data, he says, it is possible to create models that will predict criminal behavior with a fair degree of accuracy.
The basic idea is that criminals are predictable creatures. If there is an opportunity—a streetlight out on a certain corner, say—a criminal will exploit that opportunity as long as it exists. A burglar sees a dark street corner as an opportunity to break into a house, and when that crime is reported, the relevant information (where, what time, what kind of crime) feeds into the predictive policing algorithm.
Maybe that burglar decides to make another trip to the same house, or perhaps a second burglar sees the same streetlight as an opportunity to burglarize the house next door—these are the ways that one crime can foreshadow another.
“We’re not nearly as complex as we think we are,” Brantingham says.