Search

Social Networks Predict Victims of Gun Violence

A homicide crime scene. Picture courtesy of Homicide Training.
A homicide crime scene. Picture courtesy of Homicide Training.

More than 14,000 cases of homicide occurred in the United States in 2011, with over 65 percent of them caused by firearms and over 75 percent with victims that were male. It is normally presumed that homicide victims are associated with certain demographic groups, namely young African American men who are gang members. However, while such individuals are at high risk of being homicide victims, decades of social science research have proven futile in explaining why one individual in these high-risk populations becomes a homicide victim while another does not.

A recent study by Andrew V. Papachristos and Christopher Wildeman, Yale Associate Professors of Sociology, has offered a new approach to understanding this issue. Using a type of social network known as a co-offending network, they were able to take into account not only the risk factors of individuals but also the relationships between individuals in the community.  Armed with this method, they hypothesized that the closer someone is socially to a homicide victim, the more likely they themselves are to be victimized.

The city of Chicago, Illinois, from where data for this study was drawn. Picture courtesy of Howard University.
The city of Chicago, Illinois, from where data for this study was drawn. Picture courtesy of Howard University.

Using a high-crime African American community in Chicago as a model for the study, the researchers established a co-offending network that tracked the instances in which individuals were arrested together for the same crime. The use of a co-offending network assumes that people who are arrested together both participate in risky behaviors and share some sort of social tie. Based on the number of arrests over the five-year study period, the co-offending network included 24,110 individuals, or roughly 30 percent of the total community population.

By analyzing homicides in the context of this co-offending network, the researchers found that 41 percent of all homicide victims in the community were located within the network. They also found that an individual who is a part of the network’s largest component is 30 times more likely to become the next homicide victim compared to someone outside the network. Together, these results confirm the hypothesis that the further away from a homicide victim an individual is socially, the less chance he or she has of becoming the next victim.

Social relationships between different groups of individuals. Picture courtesy of the Ingenecist Project.
Social relationships between different groups of individuals. Picture courtesy of the Ingenecist Project.

Papachristos and Wildeman’s idea of using a social network as a statistical modeling strategy provides a new insight for investigating  crime victimization. By introducing another layer of information, social network analysis can more accurately predict where in the community homicide incidents take place and who is involved. Therefore, for those living in high-risk communities, who you know could be just as dangerous as what you do.