Unmasking Banksy: Did 'predictive policing' tool catch an artist?

Scientists claim to have identified the elusive graffiti artist using mapping technology and the theory that people tend to commit crimes close to their homes.

A jogger stops to take photos on his phone of a new artwork by British artist Banksy opposite the French Embassy, in London.

Alastair Grant/AP/File

March 5, 2016

A geographic profiling tool, developed to find serial criminals and terrorists, may have helped unmask the mystery identity of Banksy.

Researchers say they have identified the elusive artist – creator of million-dollar works of political graffiti – as Robin Gunningham, supporting a theory published by Daily Mail in 2008.

Scientists at Queen Mary University of London used a statistical tool to map 140 locations of Banksy's works around Bristol and London and compare them to the homes of possible candidates, they wrote in the Journal of Spatial Science. That led them to Mr. Gunningham.

In Kentucky, the oldest Black independent library is still making history

This mathematical method of analysis, known as criminal and geographic profiling, is often used by law enforcement to identify serial criminals. The idea behind the technique is that people tend to commit crimes close to where they live.

The technique has also been used to trace breeding sites for malaria outbreaks or to locate the roosts of wild bats, and the researchers suggested that what helped find one graffiti artist could also help locate terrorists.

"More broadly, these results support previous suggestions that analysis of minor terrorism-related acts (e.g., graffiti) could be used to help locate terrorist bases before more serious incidents occur," they wrote in their abstract.

Not everyone accepts that geographical profiling can accurately pinpoint perpetrators, though it’s used by several US police departments.

Data-fueled analytics also called “predictive policing,” has drawn considerable critics, arguing that the method is discriminatory and often targets minorities.

A majority of Americans no longer trust the Supreme Court. Can it rebuild?

“What data are they using? How are they weighing variables? What values and biases are coded into them? writes Jathan Sadowski, for the Guardian. “Even the companies that develop them can’t answer all those questions, and what they do know can’t be divulged because of trade secrets.”

"Police departments are opening the way for corporations to have disproportionate influence over what policing means in society," he adds. "Technologies are not just neutral tools, and they are not divorced from politics; they are designed with certain values and goals in mind."

In addition, it's too broad a tool to dependably pinpoint a criminal, argued Kami Chavis Simmons, a former prosecutor who teaches criminal justice at Wake Forest University School of Law. 

"This technology creates the risk that police view everyone in the ‘hot spot,’ even the law-abiding residents, as a potential threat," she wrote in an op-ed last fall. "This phenomenon creates tension and further destabilizes an area most in need of police protection."

In February, the Chicago police department (CPD) came under scrutiny when officers pre-emptively visited residents on a computer-generated “heat list” which marked them as likely to be involved in a future violent crime. The residents had done nothing wrong, but the CPD wanted to let them know that officers would be keeping tabs on them, reported the ACLU.

Advocates argue that the method can be useful in tracing the patterns of gun violence, as Yale sociologist Andrew Papachristos wrote on the Times blog.

Victimization is usually highly concentrated within identifiable and relatively small social networks. In one study, for example, my colleagues and I found that 70 percent of all gunshot injuries in Chicago between 2006 and 2012 occurred in networks that made up just 6 percent of the city’s population. This violence spreads like an infection among individuals as they engage in risky behaviors.

He notes, however, that interventions should have "a victim-centered public health approach in mind — one based on risk assessment and observation, rather than prediction."

In other words, use the tool to discover who to help, not who to arrest.