The friends you keep can predict your risk of gun violence

"Generally, you can’t catch a bullet from just anyone," says Andrew Papachristos. "Your relationship with the people involved matters. It’s not unlike needle sharing or unprotected sex in the spread of HIV." (Credit: John Spade/Flickr)

A person’s social network is a better predictor than race, age, gender, poverty, or gang affiliation of whether they will become a victim of gun homicide.

“Risk factors like race and poverty are not the predictors they have been assumed to be,” says Andrew Papachristos, associate professor of sociology at Yale University. “It’s who you hang out with that gets you into trouble. It’s tragic, but not random.”


Published in the American Journal of Public Health, the research likens gun violence to a blood-borne pathogen in that crime, like a disease, follows certain patterns.

People in the same social network are more likely to engage in similar risky behaviors—like carrying a firearm or taking part in criminal activities—which increases the probability of victimization.

“Generally, you can’t catch a bullet from just anyone,” Papachristos says. “Your relationship with the people involved matters. It’s not unlike needle sharing or unprotected sex in the spread of HIV.”

For the study, researchers examined police and gun homicide records from 2006 to 2011 for residents living within a six-square mile area that had some of the highest rates for homicide in Chicago.

They found that 6 percent of the population was involved in 70 percent of the murders, and that nearly all of those in the 6 percent already had some contact with the criminal justice or public health systems.

Risky networks

In addition, those in the 6 percent had a 900 percent increased risk of becoming a victim of gun homicide—suggesting that being part of a risky network might offer more insight into one’s chance of becoming a victim than other risk factors.

“You could easily identify who the dots are on these network maps and direct the resources accordingly,” Papachristos says.

Which is exactly what the Chicago Police Department (CPD) has done. In a new violence prevention initiative, the CPD is currently using network analysis to identify the area’s top 20 residents most likely to shoot someone or be shot.

A CPD police commander and the head of a community anti-violence program recently visited these high-risk individuals at their homes, explained how their history and networks landed them on the list, and delivered the message that their lives matter and they want them to stay out of trouble.

“The CPD is using this as a way to reach out to people, rather than just make arrests,” says Papachristos.

Better policing

While it is too early to know the approach’s long-term effects, other cities are taking notice and working on similar network models, including East Paolo Alto and Stockton in California, and Bridgeport and New Haven in Connecticut.

Papchristos has scaled up his study to the entire city of Chicago, a network of more than 170,000 individuals. He is currently expanding into other cities and also following the physical guns in social networks.

“Ultimately, we want to answer the question of how can we police better, smarter, and fairer,” he says.

Christopher Wildeman, was a co-author on the study, which was funded by the Robert Wood Johnson Foundation.

Source: Yale University