Is it possible to predict who is most likely to die at the hands of a gun? Not shootings like those at Sandy Hook Elementary School in Connecticut, the movie theater in Aurora, Colo., or the Washington Navy Yard, but the all-too-common shootings that occur in neighborhoods across the country.
The idea is not far-fetched if one drills down into the nature of gun violence, which, in the way it is transmitted, bears striking similarities to public health epidemics such as cholera in Haiti or HIV/AIDS in the United States.
Epidemics of any kind are not random. HIV is a blood-borne pathogen transmitted primarily through sex or intravenous drug use. Once its patterns of transmission and the communities most affected were identified in the United States, extensive public health campaigns helped transform AIDS from an always-fatal disease into one that is chronic and manageable for many of those infected.
Along the same lines, a cholera epidemic exploded in Haiti in 2010 when sewage from a base housing United Nations peacekeepers leaked into a nearby river that supplied drinking water. The transmission, far from random, ultimately killed more than 8,000 Haitians.
So it is with gun violence. There are patterns of transmission in the United States that go beyond aggregate factors such as race, age, gender and income. On an individual level, social networks - the people one hangs out with - can predict a given person's likelihood of being shot and killed.
In a study published last month in the American Journal of Public Health, my colleague Christopher Wildeman and I applied the science of social networks to patterns of gun homicide in Chicago. The idea is straightforward: Treat gun homicide like a blood-borne pathogen, something transmitted from person to person through specific risky behaviors. Put another way, gun violence is not an airborne pathogen: You don't catch a bullet like you catch a cold.
More than 40 percent of all gun homicides in the study occurred within a network of 3,100 people, roughly 4 percent of the community's population. Simply being among the 4 percent increased a person's odds of being killed by a gun by 900 percent.
These numbers tell us that gun violence spreads like HIV infection: You're more likely to "catch" the disease if you engage in risky behaviors with someone who might be infected. And it's not just people's friends who affect their likelihood of getting shot, but also their friends' friends. This is similar to the transmission of HIV: Your current partner's past sexual partners affect your exposure, even if you don't know them.
In the case of gun homicide, seemingly random victims end up "in the wrong place at the wrong time" by indirect exposure, such as getting a ride from a friend's cousin or by going to the party of a friend's friend. In these cases, victimization is tragic but not random.
Understanding the networked nature of gun violence has important implications for how it can be addressed. Prevention efforts can be directed toward those individuals and communities most susceptible to the infection. The solution is not broad, sweeping policies, such as New York's "stop and frisk" or mass arrests, but the opposite: highly targeted efforts to reach specific people in specific places, akin to providing clean needles to drug users to prevent the spread of HIV.
By studying gun violence like we study disease, we, as a society, can improve our chances of discovering who has a greater chance of being shot and focus resources to police better, smarter and more fairly.
The writer is an associate professor of sociology at Yale University.