How the Cutting Edge in Crowdsourced Crime Fighting Could Do More Harm Than Good
People have been crowdsourcing crime-fighting ever since the advent of the wanted ad. You know the kind: Here’s a blurry face, or maybe a sketch artist’s rendering of one. Have you seen this guy recently? Police are looking for him.
There’s nothing terribly offensive about such posters, which are more synonymous with the Old West than Big Brother. But what if you take photos of thousands of suspects – not just society’s worst offenders, but its pick-pocketers and purse thieves – and upload them to millions of smart phones? Police in London are actually doing this now, with an app called “Facewatch” that pushes the concept of crowdsourced criminal justice into some creepy new territory.
"My hope is that the two-thirds of Londoners who own smart phones will download this app, and help us identify those suspects we still need to speak to," a London Metropolitan Police Services official crowed in a recent press release. "We need Londoners to browse through the app every week or so as new images will appear regularly, this is a fantastic way for Londoners to help us to fight crime."
In practice, though, this could be something quite different: an invitation for people to move about the city randomly fingering each other as criminals.
"The central problem here is that people are really bad at doing eyewitness identifications," says John Roman, a senior fellow in the Urban Institute’s Justice Policy Center who has had his eye on Facewatch. "There’s a huge body of scholarship about this, and it says that people get it wrong a lot. And if the person they’re identifying is a different race than they are, they’re really, really bad at it."
Most of that research refers to scenarios where someone who has witnessed a crime in person is later trying to pick out a suspect in person. The Facewatch scenario is even more dubious, inviting people to identify suspects they’ve never seen before using a still shot from closed-circuit television (the MPS’ priority right now is to find suspects filmed during the London riots a year ago). Inevitably, people will get this wrong, Roman says, and actual law enforcement will waste time tracking down endless bad leads.
"I think the intent is to say 'that’s Tim from down the street, I recognize that guy. That’s the neighbor’s kid,'" Roman says. "The problem of course is that some people will use it that way. And some people will turn it into a game."
Think about it. Under what circumstance would you say to yourself, "I’ll think I’ll pull out my phone and troll through photos of suspected criminals"? Roman envisions that you’re probably sitting at a pub with your friends, comparing those blurry mugs to every barfly around you.
This isn’t efficient crime-fighting. It also isn’t a great way to go through life interacting with other people.
"You’re creating an environment where the social norm is to be suspicious," Roman says. "A little healthy suspicion is good. Don’t go through dark alleys late at night listening to your iPod. But to travel through your day worrying that everybody around you is a wanted criminal is pretty unhealthy."
You might argue that a similar concept works for America’s Most Wanted. But the TV show funnels massive resources at a few of the most dangerous criminals in society. The idea just doesn’t translate, Roman says, to thousands of petty offenders at a time. Especially not when smart-phone users are one of the demographics least likely to have casual contact with petty criminals.
This idea was obviously born out of the confusing modern marriage between crime-fighting and technology. In many ways, technology will make police work more efficient and effective. But it's just as likely to lure us toward downright weird ideas. Fighting crime is often about making connections. And it makes more sense to connect what police know about a given crime to other data points than to connect a photo from that crime scene to a million smart-phones users. The first solution calls for smarter IT, not apps.
London’s enthusiasm suggests police there think Facewatch could at least serve as a deterrent: If you steal stuff, your face will wind up on a million strangers' iPhones. And then everyone will be looking out for you!
"But if you care about things like civil liberties, that’s a little troubling," Roman says. "If you care about creating positive norms in society, that’s pretty troubling. Even if you just care about the bottom line, the cost effectiveness, it’s absurd."