(Illustration by iStock/Vitalii Gulenok)
Megan Kang’s research began with a deceptively simple question: Why do some people resort to violence while others do not?
Conventional explanations emphasize structural factors such as poverty, segregation, and limited economic opportunity. These forces, however important, fail to tell the entire story. Neighborhoods facing similar levels of disadvantage can have vastly different rates of violence, and even on the most violent blocks in America, most residents never engage in violence.
Kang, an assistant professor of sociology at Johns Hopkins University’s School of Government and Policy, teamed up with Kathryn Edin, a professor of sociology and public affairs at Princeton University’s School of Public and International Affairs; Timothy Nelson, a lecturer in sociology there; Jens Ludwig, a professor of economics at the University of Chicago; and Sendhil Mullainathan, a professor at MIT in the departments of economics and of electrical engineering and computer science, to look beyond structural factors and study how people perceive, interpret, and respond to situations they encounter.
The program they studied, READI Chicago (Rapid Employment and Development Initiative), brought together many different partners to help men at high risk of violence with cognitive behavioral therapy and other interventions. “It was shaped by unlikely collaboration between the Chicago Sports Alliance, business leaders, philanthropy, city officials, community organizations, and academics,” Kang says.
To evaluate READI Chicago, the researchers drew on 99 in-depth interviews with men from Chicago neighborhoods heavily affected by violence who were current or past READI participants. They used open-ended questions anchored in specific moments of conflict, asking participants to walk through events and reflect on what they were thinking and feeling to uncover cognitive processes that often work below the level of conscious awareness.
Rather than focusing on what these men had done, the researchers examined how they interpreted the social situations that preceded violence. Their interviews revealed three recurring cognitive patterns that participants identified as central to their involvement in violence.
The first pattern is what the researchers call “misconstrual”—the tendency to interpret ambiguous social cues as threats. For men who grew up surrounded by violence, this response was not irrational. When Sidney, 27, was shot seven times in 2018, his worldview shifted. “If I feel like you gonna hurt me,” he said, “I’m gonna hurt you first.” Threat-focused interpretations were often adaptive responses to dangerous environments.
The second pattern compounded the first. Participants described relying on a narrow, often aggressive set of responses to conflict. Kobe, 24, recalled that after leaving prison, he was “a ticking time bomb.” Others traced these behavioral scripts to childhood. Tony, 41, who grew up watching his father beat his mother, found himself reproducing the same patterns. Many participants had come to see themselves not as people who had committed violent acts, but as inherently violent people.
What makes READI unusual is the hypothesis embedded in its design: that behavioral patterns, while deeply ingrained, can change. The program’s cognitive behavioral component, “Control-Alt-Delete” (CAD), provided participants with structured tools to interrupt automatic responses. In sessions held three to five times a week, they role-played tense scenarios, practiced perspective-taking, and learned to recognize physiological cues—sweaty palms, a tightening jaw—that signaled an escalating reaction.
Across the three cognitive patterns, the program created space for participants to reinterpret a situation before reacting and consider alternative responses. Seeing past behavior as shaped by circumstance, not fixed character, they separated what they had done from who they were.
“This is a novel and actionable contribution, made possible by the authors’ sophisticated integration of behavioral economics and sociology,” says Robert Sampson, a professor of sociology at Harvard University.
Kang notes that the study brings together disciplines that rarely intersect: “Psychologists focus on cognitive processes; sociologists focus on the conditions that produce them. But what if conditions shape cognition, and cognition drives behavior?”
The results of READI’s randomized controlled trial suggest its approach has impact. Participants showed a 65 percent reduction in arrests for shootings and homicides, with estimated social benefits of up to $916,000 per participant.
Find the full study: “Social Cognition and Interpersonal Violence” by Megan Kang, Kathryn Edin, Jens Ludwig, Timothy Nelson, and Sendhil Mullainathan, Proceedings of the National Academy of Sciences, vol. 123, no. 19, 2026.
Read more stories by Daniela Blei.