The San Francisco district attorney’s office is just around the corner from some of the world’s most cutting-edge data collectors and data miners. Yet, according to DA George Gascon, his office is stuck in the 1970s. To leap forward a few decades, Gascon has hired his first chief information officer.
Now he’s got to warm his prosecutors to the idea that cold hard data can help deliver justice.
Gascon wants to give his office a makeover (and not just with furniture), using Big Data. He says the data Revolution that’s transformed whole industries has been totally lost on the criminal justice system. For instance, in response to a charge that Vietnamese and Korean defendants are often misclassified as Chinese, he says, “Yeah, I think it’s even worse. In many cases they only are classifying white, black and other.” Regarding another allegation, that his office is letting domestic violence perpetrators off the hook, he says, “Is this accurate? I cannot come back to the office and push a button and get a report that is going to tell me this.”
He plans to change that with an initiative he calls DA Stat.
The DA oversees 130 prosecutors and a $42.7 million annual budget. Now, in addition to money for the new chief information officer, the city is doling out $320,000 for a new database to replace the current relic.
“The database was very, very old,” Gascon says. “I think it might not quite be the ’50s. But I think it’s sort of ’60s, early ’70s vintage and … most of it was handmade, I think.”
That database, called DAMION, does track conviction rates, the metric historically used to evaluate prosecutors. But Gascon also wants to track demographics, crime categories, punishment and reoffense rates — something to figure out the odds, statistically, of someone reoffending. He hopes to be able to do that with this new tool.
Take someone charged with theft. In terms of both likelihood to reoffend and the most effective type of intervention,” a first-time offender age 45 is going to be very different from a first-time offender age 16,” Gascon explains. It’s not enough for a prosecutor to rely on gut instincts when deciding whether to go light or throw the book, he says, because “when you do it intuitively, we end up incarcerating and making the wrong decisions too often.”
Individuals, not statistics
This concept – that to stop defendants from becoming statistics, you’ve got to turn them into statistics – is not without controversy.
California Attorney General Kamala Harris used to have Gascon’s job. In her book, “Smart on Crime,” Harris advocates a metrics-driven approach to punishment and community programs, but does not go as far as Gascon in advocating for Big Data.
Some lawyers say criminal justice shouldn’t try to mimic the insurance industry, in which pre-existing conditions can work against you, for instance. The 16-year-old thief who’s statistically more likely to reoffend may have been simply reacting to a one-time event, stealing to pay bills after the death of a parent, for example.
Fanya Young, a former prosecutor turned defense lawyer, is skeptical of DA Stat. She says that when you turn people into statistics, you risk losing their personal stories.
“There’s not really a connect between statistics and prosecuting or defending a case,” she says. “A case is a case. A person’s individual rights are a person’s individual rights.”
Holding DAs accountable
There’s another reason prosecutors might resist. The new approach will leave them open to new scrutiny.
Gascon copped the name DA Stat from his old boss, William Bratton– the police chief renowned (or reviled) for creating COMSTAT, the data system Bratton implemented back in the 1990s in New York City to map out crime hot spots and deploy police resources accordingly. This shook up some NYPD personnel practices; for instance, middle managers had to account for time. Bratton explains, “There’s no place to hide in CompStat. It’s intended to ensure that things are not hidden.”
Bratton says DA Stat can do for prosecutors what CompStat did for police. “Take a look at the various assistant district attorneys, what their caseload is, what their conviction rate is, the length of time it takes to try a case. There is no shortage of information in a district attorney’s office that can’t be CompStat’ed.”
Bratton says his protégé is the perfect guy to lead the change, because unlike most prosecutors, Gascon used to be a cop. He saw Compstat work firsthand in Los Angeles.
The secrets of prosecutors
Joan Petersilia, a law professor at Stanford University, says Gascon will have a tough time selling DA Stat because of the elusive nature of prosecutors.
“It’s always said about prosecutors they have the most power and we know the least about what they do and how they exercise that power,” Petersilia says. “They’ve always been the hidden part of the criminal justice system.”
Back in the 1980s, Petersilia advocated that prosecutors use data to figure out why so few cases led to conviction. Practitioners didn’t know if the problem was a glut of cases brought without proper evidence or a lack of resources to pursue strong leads. She helped roll out a database called PROMIS, Prosecutor’s Management Information System. DAs nationwide bought in, but then kept the data to themselves.
Petersilia says instead of creating transparency, PROMIS simply became an internal management tool. She’s urging Gascon to learn this lesson well. “Sharing that information to the probation department, to the public defender’s office, to community-based organizations, he will actually be evaluating the impact of the criminal justice system in San Francisco.”
Gascon knows DA Stat won’t be easy to implement. He says it’s worth the heavy lift because “the criminal justice system is a product at the end of the day, just like when you buy cars or you buy a home. And I think that if we want to be good consumers of public safety, then we need to be informed.”
Gascon just hired his very first CIO. Now he’s hoping he can get a database that’s not of the “vintage” or “handmade” variety.
Listen to the radio version of this story: