After experiencing a homicide rate that earned it international attention last year, Chicago is upending the traditional style of policing and using social networks to rank people’s likelihood of killing and being killed.
It’s Friday afternoon on the far West Side of Chicago, and Barbara West, a petite, soft-spoken African-American woman, is out knocking on doors. West isn’t a salesman or solicitor; she’s the commander of the Chicago Police Department’s 15th District. Today, she is visiting a special subset of her constituents—the 20-plus people who are likeliest to shoot someone or be shot themselves. She’s accompanied by Chris Mallette, a 6-foot, 2-inch, 250-pound former linebacker and football coach who now heads the Chicago Violence Reduction Strategy, an anti-violence initiative funded by the MacArthur Foundation and supported by the John Jay College of Criminal Justice in New York.
The 15th District—like Chicago as a whole—has a murder problem. In cities such as New York and Los Angeles, homicides have fallen by nearly 80 percent since the early 1990s. Crime in Chicago has fallen too, but its murder rate has remained stubbornly high. While the homicide rate today is only half of what it was a generation ago, Chicago residents are twice as likely to be killed as New Yorkers or Angelenos. For young African-American men that likelihood is higher still. One out of every 400 young black men is killed each year in Chicago’s highest-crime neighborhoods. Much of the violence is associated with the city’s gangs. Chicago Police Department officials estimate that 50 to 80 percent of the city’s shootings and murders are gang-related.
It isn’t just Chicago. Some 80 percent of the nation’s largest cities and half the country’s suburbs report significant gang problems. But Chicago’s gangs are different. First, there’s the scale of the problem. With some 100,000 documented gang members, Chicago has more gang members than any other city with the possible exception of Los Angeles. (Researchers estimate that gangs account for at least half the homicides in those two cities, a number so large that together they make up about 20 percent of all gang-related homicides nationwide.)
Last year, Chicago experienced a spike in homicides that brought national—indeed international—media attention. The 15th District’s murder rate surged in 2012, due in part to a feud that broke out within one of the area’s largest gangs, the Four Corner Hustlers, a faction of the Gangster Disciples that controls the area’s lucrative marijuana trade. Most police departments would have responded by “flooding the zone” with additional officers, a tactic known as “hot spot” policing, and perhaps targeting the Four Corner Hustlers for narcotics operations. In fact, the Chicago Police Department (CPD) has done just that. But it’s also doing something far more unusual. It’s figuring out exactly who is likeliest to kill and be killed in each district.
The 20-some people that West and Mallette are visiting all have connections to the neighborhood’s 48 active gang factions. But what makes these individuals extreme risks is whom they associate with—in other words, their networks.
We live in an age of social media, where companies such as Facebook and LinkedIn assiduously map our friendships and use that knowledge to shape our preferences and behavior. Today, the Chicago Police Department is doing something similar with gangs. Using a tool academics call “network analysis,” the CPD is mapping the relationships among Chicago’s 14,000 most active gang members. It’s also ranking how likely those people are to be involved in a homicide, either as victims or offenders. In the process, the CPD has discovered something striking: Cities don’t so much have “hot spots” as “hot people.” That finding is transforming the way the police do business in Chicago and has significant implications for how other cities should be policed.
The idea is closely connected to the department’s enthusiastic embrace of network or link analysis, an embrace that began with the work of an unorthodox Chicago-born sociologist, Andrew Papachristos.
For Papachristos, Chicago’s gangs were not an abstraction. He grew up in the 1980s on the city’s North Side, in Rogers Park, one of the few neighborhoods in Chicago with roughly equal proportions of black, Hispanic and white residents. His father, a Greek immigrant, ran a neighborhood diner. His mother worked with the Guardian Angels (as Papachristos would years later), a group of volunteers that assisted police in patrolling high-crime neighborhoods, and she ran a small nonprofit that provided kids with after-school activities and jobs. It was the height of the crack epidemic, but gangs didn’t limit their activities to drug dealing. Extortion was also a favorite activity. Diners like the one operated by the Papachristos family were a favorite target. Papachristos’ father refused to pay the neighborhood tax. And worse, his mother’s work with the Guardian Angels helped get kids off the street. Some kids got jobs washing dishes at the diner. Others sat in the back of the restaurant and did their homework. That made the family a target.
“They warned us,” says Papachristos of the local neighborhood gang. “Eventually they torched the place.”
Not surprisingly, Papachristos thought about becoming a cop. But as a student at Loyola University, he took a course in criminology. He was immediately intrigued. Although he had a job offer from the police department in suburban Oak Park, a former high school teacher encouraged him to instead get a Ph.D. in sociology. At the University of Chicago, he began to conduct ethnographic studies on the West Side and came to the conclusion that attempts to explain urban violence in terms of factors such as poverty and race did a disservice to many of the people in the highest-crime communities. Most residents of even the poorest neighborhoods didn’t kill or commit violent crimes. Most gang members didn’t either. But when they did, it was often as members of a group.
“These groups have agency. They have structure. And it’s not random. In fact, the big thing there was that they’re inheritable,” says Papachristos. That is, gangs don’t necessarily know why, generation after generation, they shoot one another. Papachristos likens it to why he “hates” Green Bay Packer fans—something he and his neighbors do and his parents and their neighbors did. “I don’t actually hate Packers fans, but that sort of structure actually dictates what I wear, who I watch football with, who I tease on Facebook. That’s the same sort of things that gang members inherit, and it leads to murder.”
Sociologists and anthropologists have long talked about murder as a “gift,” something to be “given” with the full expection that it will be reciprocated. Papachristos wanted to know exactly how giving the gift of murder worked, who “gave” it to whom. Boston had done something similar in the early 1990s. In response to a surge in youth violence, an interagency group there mapped gang feuds and identified the most active players. The group then called those gang members in for meetings with police, probation officers, prosecutors, gang workers and social service agency representatives to deliver a message: “We know who you are. We want to help you stop shooting. If you don’t, we’re coming after you.”
The initiative, known as Operation Ceasefire, dramatically reduced violence and inspired widespread interest in an approach to crime reduction known as “focused deterrence.” Papachristos wanted to do something similar using arrest records and contact cards, which record police encounters with citizens. When an arrest occurs, the police generate a unique event ID. That ID includes the names of the people arrested. Papachristos assumed that people arrested or stopped together were linked. After all, he says, “You don’t rob a bank with a stranger.” He hypothesized that many of the people killed would appear in the network. By seeing who was connected to whom, police could better understand who might retaliate and who was at risk.
In 2008, Papachristos began experimenting with ways to take this idea and make it operational. Working with data from a big-city police department in the Northeast, he demonstrated that charting networks of homicide victims not only successfully identified individuals that police were already aware of; it also identified people connected to gangs in more surprising ways. That work captured the attention of the Chicago police. Soon Papachristos was doing similar analyses for the 11th and 15th police districts. What he found was striking. People who were friends or friends of friends with homicide victims were approximately 100 times more likely to be involved in a future homicide than people who weren’t. Moreover, the people within this network appeared as future homicide victims and offenders at roughly equal rates. They weren’t perpetrators or victims. They were both.
The implications of these findings were intriguing. For generations, the Chicago Police Department, like most police departments, had thought about certain areas as dangerous. The area from which Papachristos was pulling data—the Harrison District of North Lawndale—was one of those dangerous areas, or hot spots. The homicide rate for Chicago as a whole at that time was 14.5 per 100,000 residents. In North Lawndale, it was significantly higher—44.5 per 100,000. But when crime was considered in the context of networks rather than places, the picture changed in interesting ways. Papachristos found that people who had been arrested within the previous five years were 50 percent more likely to be killed than people who were not arrested during that period. Even more interesting were the heightened risks faced by people within two degrees of association with a homicide victim: The homicide rate for that group was a shocking 554 per 100,000, or 900 percent higher than the average Chicagoan. For people who had never been arrested, the homicide rate fell to just 15 per 100,000, a rate virtually identical to that of the city as a whole.
Papachristos continued to work with the police from the 11th and 15th districts. However, it wasn’t until 2011, when newly elected Chicago Mayor Rahm Emanuel announced the hiring of a new police superintendent—Garry McCarthy, an alumnus of the New York City Police Department (NYPD)—that network analysis became one of Chicago’s primary tools in responding to gang violence.
McCarthy grew up in the Pelham Bay area of the Bronx, the youngest of three kids. His father was an NYPD detective; his mother, a nurse. McCarthy joined the department in 1981. He rose quickly, making sergeant in 1985 and captain in 1992. McCarthy began to make his mark after then-Mayor Rudolph Giuliani tapped Bill Bratton to head the NYPD. Bratton and his successors believed that precinct captains should be given the resources to address crime in their districts and then be held accountable for results, using CompStat, the department’s computerized crime tracking and accountability system. Officers who failed to accept the new rules were transferred or sidelined. Officers who thrived were promoted. McCarthy thrived. By the mid-1990s, he was running CompStat; by the time of the 9/11 attacks, he was the department’s deputy commissioner for operations with responsibility for crime control strategies.
In 2006, Newark Mayor Cory Booker lured McCarthy across the Hudson River. McCarthy brought CompStat with him, as well as a willingness to innovate rapidly. When a wave of carjackings hit the city, McCarthy deployed plainclothes police academy cadets to the hot zones. McCarthy’s sometimes-brusque style occasionally rubbed Newark City Council members the wrong way, but he maintained Booker’s support by presiding over a significant drop in violent crime even as his department was forced to absorb budget cuts.
Emanuel, himself a public figure with a hard-charging, get-things-done style, saw in McCarthy a kindred spirit. He needed a tough, smart guy to not only reduce crime in Chicago, but also to focus on police-community relations.
In the years following his departure from New York, the NYPD had relied increasingly on a tactic known as stop, frisk and question. By flooding high-crime neighborhoods with police officers and encouraging those officers to interact with residents, the department hoped to deter young men from carrying guns. By making it riskier to carry guns, the police hoped to prevent disputes from escalating to shootings. In Chicago, McCarthy’s predecessors had used two large task forces to flood hot spots. Citywide units were also frequently used to target drug dealing.
But such tactics came with a price. Because most crime occurred in majority black and Latino neighborhoods, huge numbers of young minority men were being stopped by police. The sheer scope of the stops raised questions about whether residents’ constitutional rights were being respected. (In August of this year, New York City Judge Shira Scheindlin ruled they were not, a finding the department is appealing.) Even if the stops were legal, however, many residents believed they were not legitimate. A growing body of research, much of it conducted by Yale University psychology professor Tom Tyler and Yale Law School professor Tracey Meares, suggested that perceptions of fairness are critically important to ensuring compliance.
For these reasons, McCarthy wanted to repair relations with residents. He disbanded the specialized units his predecessors had used to flood crime hot spots and reorganized patrols so that officers spent more time on their beats, another step intended to build knowledge and cooperation. Crime in Chicago fell by 10 percent in McCarthy’s first year in office. In early 2012, however, CompStat revealed something disturbing. While overall crime was falling, homicide was spiking. Moreover, most of the violence was associated with something McCarthy and other members of his N.Y.-centric team had relatively little experience with—gangs.
“In New York, we have Bloods and Crips, but they’re really narcotics gangs,” says McCarthy. “They don’t shoot each other [just] because they’re on another team.” But they were shooting each other in Chicago, which was “a very substantial difference from what I was accustomed to,” McCarthy says.
Gangs in Chicago were involved in the drug trade too, but the dynamic in Chicago was quite different. In many parts of the city, gangs were changing in ways that increased the number of conflicts. During the 1990s, Chicago began to dismantle the massive housing projects that were the home base of Chicago’s most powerful gang, the Gangster Disciples. At the same time, federal prosecutors also went after the leadership of the gang. The result was a proliferation of new smaller gangs that were nominally Gangster Disciples but that frequently feuded with one another. The disputes that resulted were often settled with guns.
Guns were Chicago’s other problem. New York state’s restrictive gun laws gave police officers numerous tools to use against people who sought or used firearms. California has similarly restrictive gun laws. In both states, people apprehended carrying a concealed firearm illegally face severe punishments. Not so in Illinois, and in Chicago, there was easy access to firearms.
“We seize more guns than any city in the country every single year,” McCarthy says. “That’s because of the structure of the gun laws in the state of Illinois, which we’re trying to get fixed.”
Faced with this upsurge in violence, McCarthy turned first to the NYPD playbook. He launched Operation Impact, which flooded 20 high-crime areas of Chicago with police officers. It was a classic example of hot spot policing. But McCarthy also wanted to understand what was happening at a more granular level. To get a comprehensive overview of the city’s gang situation, he commissioned a gang audit. In addition to documenting that Chicago had 59 main gangs, all of them familiar to law enforcement authorities, the audit also found that the city had 630 gang factions, about a hundred more factions than police had previously believed they were dealing with. “That was driving up the violence,” says Deputy Chief Robert Tracy, “[the factions] going back and forth with each other.”
As one of the strategies for addressing this violence, McCarthy launched an initiative modeled after Boston’s Operation Ceasefire, which relied on calling in gang members to deliver the “we know who you are” message. But call-ins suffer from a problem. Only a small percentage of gang members regularly resort to violence. Police call these individuals high-impact players. But the only gang members who can be required to attend a call-in are those under court supervision. As a result, authorities are sometimes unable to deliver the message to the person who most needs to hear it. No message, no deterrent.
High Point, N.C., of all places, figured out how to get around this problem. First, the police identified the high-impact players. Then law enforcement worked with prosecutors to draw up a customized letter, telling the recipients exactly what would happen to them if they were involved in a future act of violence. Since High Point began its custom notification process five years ago, the city has seen violent crime fall sharply.
McCarthy wanted to do something similar in Chicago. The scope of the challenge, of course, was much greater. A preliminary analysis indicated that it had 14,000 “hot people.” However, it also had greater resources.
Once McCarthy embraced the idea of network analysis, the Chicago Police Department moved quickly to flesh out the idea. They started by revisiting two years’ worth of murder victims. Police then looked for instances where those victims had been arrested and whom they’d been arrested with. Working with Papachristos and Illinois Institute of Technology professor Miles Wernick, the department built a model to help it identify Chicago’s “hottest” residents—that is, the likeliest to be involved in violence. The model includes such variables as how many times a person had been shot, how many times a person had been contact-carded, whether he had any gun convictions and whether he was on probation or parole. By the spring, they had identified a top 20 list for each of the city’s 22 police districts.
Being on the list means extra scrutiny. “Identify them,” says McCarthy. “Where appropriate, you take all enforcement action.” But McCarthy wanted to do more than bring greater scrutiny to bear on “hot people.” He wanted to prevent them from resorting to violence in the future. To do that, police officers needed to do something even more unusual. They needed to talk to them.
On a Friday in late July, West and Mallette, along with two district intelligence officers, kicked off the new approach by visiting the 15th District’s 20 “hottest” residents. Not all of them were at home, but West and Mallette were happy to talk with so-called “influentials”—parents, grandparents, aunts, uncles and girlfriends. In those conversations, they explain why people’s criminal histories and networks have put them on the department’s hot list. For many of these friends and family members, the status of their loved one comes as a surprise.
Many times the family members and loved ones have no idea how close these people are to violent crime. “They don’t know that he has been classified as an armed habitual offender,” says Mallette. “They don’t know probation is a conviction. They don’t know that the next time he could face federal prosecution, that Illinois has state [rackteering and corrupt organizations] laws now, that he has been identified as a known gang member. Then all of a sudden the family is like, ‘We don’t want that. We need to talk with you.’”
Sometimes, the people themselves don’t know. West and Mallette visited one man who’d been arrested 32 times. When they presented him with that information, he protested, insisting he’d only been convicted half a dozen times. They showed him his rap sheet. “Gambling counts?” he asked in response.
All told, only one person refused to answer the door. West and Mallette left people with letters that explain why they received a visit, what the consequences for continuing to offend will be and what opportunities exist to help them make positive changes to their lifestyle.
Just before dinner, West and Mallette knock at the door of someone on the hot list. He’s inside cooking shrimp. He invites them in and listens with evident amazement as West explains why they are visiting him. It’s her job to play the heavy, something she does well. West is a forceful presence. The man listens carefully while she explains what they know about him and why they are visiting him. If he gets arrested with his associates again, he will face different consequences.
“They could go to Cook County,” she tells him. “You face a federal prosecution where the conviction rate is 96 percent.”
Then Mallette steps in to deliver the community message. “We don’t care who you are affiliated with or what your rank is,” he says. “We don’t care how many times you’ve been arrested. What we are concerned about is you—you living. We love you; we value you; and we need you in the community to stand in your rightful place.”
It doesn’t always work. After Mallette delivered this message to the aunt of one person on the hot list, she just shook her head. “He’s just a bad seed,” she replied. Another woman, a grandmother, said she was just waiting for the knock on the door informing her that her grandson was dead. But this time at least, the guy in the kitchen seems to be open to what West and Mallette are saying. Two people he knew well have recently been shot. His network is risky.
Whether such approaches will change offenders’ behavior remains to be seen. But it’s already changing how the CPD operates. It used to be that when a shooting occurred, all the police converged at the scene of the crime. No longer. Now, they also deploy around the people and locations where network analysis suggests the next shooting is likely to occur.
“There’s a saying in baseball, ‘You’ve got to know what you’re going to do with the ball before it gets hit to you,’” says McCarthy. “If you don’t have that information in your hands before [a shooting occurs], you can’t possibly get out in front of it.”
As of late July, Chicago had experienced 76 fewer murders than it did in 2012. Victims of gunfire were down by an even greater number—350. Summer wasn’t over, but if the trends continued, Chicago’s violent crime level would fall to levels not seen since the early 1960s. And that would suggest that social-network policing is the future of crime fighting.
This article originally appeared in GOVERNING magazine.