Text by Andrew Papachristos and Charles Barber
Photographs by Paola Chapdelaine
“Tough Love” documents the efforts of violence intervention workers to interrupt cascades of gun violence in American cities. Most of the workers are formerly incarcerated persons, now recovered from criminal behavior, and are perceived as trusted messengers in their communities. They are now pairing with university data scientists to better identify and provide outreach to those most vulnerable to gun violence, based on known patterns of how shootings flow through social networks. The project begins in New Haven, a city where two distinct realities coexist, with the Connecticut Violence Intervention Program (CTVIP).
For five years, between 2016 and 2021, William Outlaw, a violence intervention worker in New Haven, did everything he could to engage Ciera “Cee Cee” Jones. Cee Cee was the only female member of a thirty-odd member gang, and she was allegedly a “shooter” in an ongoing beef with another gang across town. Outlaw helped her get a factory job, successfully encouraged her to return to high school, bought her prom dress with his own money, and attended her graduation. After she posted a rap song taunting the rival gang on social media, he implored Cee Cee to take it down. He visited the rival gang and personally asked them not to retaliate. All to no avail: Cee Cee was fatally shot in the head outside her house in broad daylight on July 5, 2021. She was the third member of her family to be shot to death: her niece and another relative were murdered in the previous year. Last year, a member of the rival gang confessed to Cee Cee’s killing as part of a plea bargain. That gang has been implicated in three murders and fifteen attempted murders since 2018.
In this case, William Outlaw was unsuccessful in helping Cee Cee exit from gang life, but his work testifies to his profound understanding of emerging scientific research that shows how social networks — the people one hangs out or associates with — can predict a given person’s likelihood of being shot and killed. Our research, which was begun in Boston but brought to fruition in Chicago, repeatedly confirms that less than 4 percent of any neighborhoods’ population is engaged in 80% percent of the violence. More alarmingly, being in that small social networkincreases a person’s odds of being killed by a gun by 900percent.
In other words, gun violence is tragic, but it is decidedly not random. Like an infectious disease, it cascades in often predictable patterns through social networks. One of our studies in Chicago, for example, has connected a single shooting to 64 subsequent shootings.Without being aware of our studies, Outlaw, a former gang leader, already knew the infectious nature of gun violence within close social networks from a life spent on the streets. He understood that the best chance to save Cee Cee’s life was to help her depart her perilous social network and engage with a new one of employment and education. Indeed, Outlaw’s group in New Haven, the Connecticut Violence Intervention and Prevention Program, spends most of its efforts replacing antisocial networks with positive, or prosocial, ones. To spend time with the team is to witness them sponsor weekly basketball tournaments, free car washes, yoga sessions for family members who have lost loved ones, community walks, educational sessions for Yale medical trainees, meetings with the Board of Aldermen about neighborhood trends and safety, listening to youth, and more.
But what if William Outlaw, with all his hard-won street savvy, was also armed with data that showed and directly quantified the extraordinary risks that Cee Cee was taking? Could that have made a difference? Could he have saved Cee Cee’s life?
That may sound like a “pie in the sky” hypothetical, but right now in Chicago, our team at Northwestern University is doing exactly that. For years, we have been assembling data on the social networks of gunshot victims and building relationships with street outreach organizations that are doing front-line work. Starting in 2021, we started the Street Analytics Outreach Response (SOAR) Initiative, which works with four street outreach organizations. By combining our expertise in network science with their expertise in local conflicts and violence reduction strategies, we believe we can create a “force multiplier” to reduce gun violence.
Weekly, the SOAR team gathers around a table with outreach teams. Researchers bring data analytics and graphs of the previous week’s shooting victims while outreach workers come equipped with their intricate neighborhood knowledge. Collaboratively, we investigate scientific diagrams of the previous week’s shootings that resemble spider webs, with lines and dots of different shapes and colors. These are network diagrams, showing the connections between recent shooting victims and their associates. Red dots indicate those shot or killed, while the connecting threads show potential relationships with those victims. Those red dots provoke discussions about who might be at risk next. Researchers walk through the graphs while outreach teams trace the connections in their heads that align within the images, looking for people they know that might provide an inroad into their participants’ lives. “That’s Mike’s cousin,” an outreach worker will say. “He used to run with a crew from 18th Street, we know that crew.” Within minutes, workers mobilize to prevent retaliation: making calls, sending texts, and conducting urgent visits.
Quite often, however, the network images reveal connections that even the most street-savvy workers couldn’t have known. When names of victims or involved parties aren’t familiar, the data process might provide breadcrumbs for outreach workers to follow. A shooting on the South Side might be linked to a crew on the West Side, or to a neighborhood where a different outreach group operates; help might be a phone call away. This expanded view allows outreach workers to navigate social terrains far beyond their usual borders, coordinating responses across the city with unprecedented precision.
It’s as if we’re giving them a social GPS, extending street smarts beyond the limitations of a given worker’s memory or personal connections. After all, an outreach worker like William Outlaw may know the ins-and-outs of particular groups within a few city blocks but have little information on an opposing group a few blocks away, let alone gang dynamics in another part of the city. Outreach teams can spend days fumbling through Facebook and their own foggy memories when they might only have hours to find someone before they pull the trigger. At the same time, while our research team may have advanced degrees and quantitative expertise, few or none of us know the generational history of a neighborhood, a gang’s internal dynamics, or who is particularly influential within a specific network. One group has what the other lacks.
After a recent shooting in Chicago, the SOAR process allowed outreach workers to link a recent victim to the murder of his brother a year earlier by tracing lines between shooting victims on one of the scientific graphs. This didn’t occur until one of the workers inquired, “Hey, who’s this dot right here?,” pointing to another red dot on the screen. But once that the familial connection was made, another outreach work linked it to something that no statistic in any database could ever reveal: there was a third brother who might also be in danger. As it turned out, that brother was in school with the son of one of the outreach workers. The team then leveraged this school-based connection to reach the grieving mother who had lost two sons. At the outreach worker’s urging, and with the outreach organization’s support, the mother and son temporarily relocated out of Chicago to allow the dispute to cool.
While SOAR is in its early days, it is already clear from cases like this that the collaboration has extended the reach of these violence prevention organizations and is opening new opportunities for effective collaboration. But the potential of bridging street smarts and data analytics goes even further. The extreme concentration of violence in networks means that bringing them into sharper focus might allow for better interventions—a fact especially important given the small scale of most CVI programs. In Chicago, a city of three million people, there are only 300 outreach workers in a city of three million, and by some estimates, 900 gang structures; in contrast, there are more than 12,000 police officers. Making sure that outreach workers can efficiently track the ebb and flow of the social dynamics of violence is vital to saving lives.
William Outlaw agrees. “I think about what else I could have done with Cee Cee most every day,” he said. “In the war that is played out every day on the streets of America, I need every tool I can get.”
