AI Identifies 160 Possible 'Crews' of Criminal Cops in Chicago

A new study has used machine learning to identify over 100 possible "crews" of officers in the Chicago Police Department responsible for a disproportionate share of police misconduct and violence. 

The study—authored by researchers at Northwestern University and the Invisible Institute and published on Wednesday in PLOS ONE—identified approximately 160 groups of officers, referred to as "crews." The identified groups consisted of less than 4 percent of Chicago's police officers but accounted for about 25 percent of "all use of force complaints, city payouts for civil and criminal litigations, and police-involved shootings," These crews generated about 18 percent of all complaints filed by Black residents and 14 percent of all Hispanic residents, according to the researchers. 

To identify cops who intentionally group themselves into crews, the researchers used data from sources including complaints, assignments, and settlements to see which officers were involved in "alleged co-misconduct," how often recurrent co-misconduct happened within the group instead of external to it, whether officers in a crew engaged in "similar types of alleged misconduct," and whether there seemed to be well-defined membership whether by age, location, race, or other attributes.

"Explanations for police misconduct often center on a narrow notion of 'problem officers,' the proverbial 'bad apples,'" the researchers write. "Such an individualistic approach not only ignores the larger systemic problems of policing but also takes for granted the group-based nature of police work. Nearly all of police work is group-based and officers' formal and informal networks can impact behavior, including misconduct.”

For years now, conservatives and law enforcement officials have clung to the narrative of “bad apples” theory of policing: that individuals, not institutions, are to blame when police misconduct and violence happen. One might ask how departments gorging themselves on taxpayer funds to acquire surveillance technology or ostensibly train investigators and detectives might miss the bad apples, but that’s ultimately because, as sociologist Rashaw Roy has put it multiple times and the new study attempts to elucidate, “bad apples come from rotten trees.”

In 2021, the Financial Times analyzed officer complaint data in Chicago, New York, and Philadelphia and found that while thousands of officers in each city received at least one complaint since 2007, ten percent of them accounted for a third of all complaints. In their analysis, the Financial Times pointed to previous research co-authored by one of the researchers involved in Wednesday's study—Northwestern sociology professor and director of the Northwestern Neighborhood & Network Initiative Andrew Papachristos—that uncovered "large network structures" of misconduct and abuse. That study identified not only abusive officers but also "clusters" whose interactions "contribute to the emergence or diffusion of misconduct” and could be considered when creating assignments, patrols, and promotions.

Applying network analyses to police departments makes sense given this history of policing and groups of officers generally. Take two prominent examples of crew-equivalent misconduct cited in the study: the Los Angeles Rampart and the Oakland Riders scandal. 

In the former example, more than 70 officers in the LAPD were involved in "assaults, drug crimes, fabricating evidence, perjury, and, allegedly, murder," the researchers write. In the latter, four Oakland officers were "accused of actions ranging from filing false reports to assault, kidnapping, and severely beating civilians with their fists, feet, pepper spray, and metal clubs." Both crews coordinated actions amongst themselves, covered up activities, and in the case of the Rampart crew, developed into a gang with its own symbols, oaths, and language.

These aren’t isolated incidents. Reporter Cerise Castle has documented similar activity and worse in Los Angeles as part of her magisterial 15-part investigative series into the Los Angeles County Sheriff's Department and its deputy gangs. For the past 50 years, according to Castle's reporting, at least 18 gangs have terrorized communities of color in the county, killed at least 19 people (all of whom were men of color), are responsible for a disproportionate of misconduct and violence, and have cost the County just over $100 million in litigate over the past three decades.

Police misconduct has a long history in Chicago, which Thursday's study focused on. The city's police department is the second-largest in the country (14,000 officers and civilian personnel), enjoys a budget exceeding $1.7 billion, and has a history of "misconduct, abuse, and corruption" that stretches back to the late 1800s, according to the study. The team of researchers also highlight how Chicago police helped assassinate the 21-year-old Black Panther chairman Fred Hampton Sr., were enthusiastically involved in COINTELPRO, regularly engaged in illegal and unconstitutional surveillance, and enjoy "a lack of effective oversight and accountability" to this day. 

Researchers point to a 2017 Justice Department report that found "repeated and frequent instances of excessive use of force, misconduct, and derogatory language and behaviors towards civilians, especially minority civilians." The same report also found the CPD took insufficient steps and often blocked efforts to stop or investigate misconduct, and had a "pervasive cover-up culture." In 2016, before the report’s publication, Mayor Rahm Emmanuel faced calls to resign as it was revealed his lawyers tried to delay the release of video footage showing Laquan Macdonald’s murder by police officers in 2014 until after the criminal trial of the officers involved.

As tempting as it may be to lay the issue at the feet of individual officers (17 percent of CPD officers generate 50 percent of all complaints, according to the study) or officers who happen to share misconduct allegations due to working together, the researchers argue it is organized groups of officers who have always coordinated, engaged in, and covered up misconduct and violence. 

"Given the group and partner-oriented nature of policing in Chicago, one of our foundational analytic tasks is to differentiate group misconduct that is a product of the organizational structure of police (e.g., two officers who receive a complaint merely as a function of working together) from those instances of more intentional crews."

In the study, researchers point to Chicago’s already  three well-known crews: the Watts Crew, The Skullcap Crew, and the Austin Seven.

The Watts Crew became an "integral part" of the drug trade in communities it policed; more than 212 convictions have been overturned because of misconduct by the crew but only two officers have been convicted. Five officers (four are still on the force) make up The Skullcap Crew, which built up 138 allegations of misconduct. Officers were regularly involved in "many reported instances of excessive and unwarranted force, sexual abuse and harassment including strip searches, planting drugs, theft, and false arrest," the researchers write. The Austin-Seven Crew was a group of seven officers "involved in cases of robbery, extortion, and drug dealing” that worked with local gang members to commit violent crimes. All seven were arrested and charged in 1995.

Still, despite the insights garnered there are limitations to the study as it doesn’t tell us how crews emerge, nor does it tell us how culpable officers are as members of crews and the larger network structure of a police organization. Finally, the researchers write, identifying organized groups of bad cops is just the start—this sort of activity has never been a mystery, and there needs to be the will to do something about it. 

"The analysis presented here demonstrates the possibility of systematically using data to identify networks within a police department that may prove, when investigated, to be criminal crews," the researchers write. "However, even the best efforts at identifying and validating such crews will only go so far without the capacity to fully investigate such cases and, when deemed necessary through due process, discipline, dismiss, or otherwise hold accountable the officers involved."

CPD did not respond to a request for comment.