Every time a civilian files a complaint about an encounter with NYPD officers, the Civilian Complaint Review Board investigates. The independent police watchdog determines which officers were accused of excessive force, improper searches, or other forms of misconduct, and which ones, if any, stood by as the interaction unfolded. Each complaint investigation thus generates invisible links tying together the small group of officers who were there that day.
Combining hundreds of thousands of these links together, a massive network emerges, revealing tightly-knit clusters of officers that civilians have repeatedly accused of abuse. With the help of independent data journalist EJ Fox, WNYC/Gothamist mapped this network using a dataset of civilian complaints obtained through a Freedom of Information Law request.
The lines on the maps above and below are unique civilian complaint co-occurrences tying together pairs of officers who have been named together in a complaint investigation, either as witnesses or as subjects of allegations. The officers themselves are represented by circles or nodes, which are sized based on the number of ties they have to other officers.
The tie count incorporates both the initial connection between two officers identified in a complaint investigation together, and each subsequent complaint investigation in which the same officer pair appears again. For example, Daniel Pantaleo, who was fired in 2019 after an investigation into his fatal chokehold of Eric Garner, had 184 network ties with 27 other officers, putting him within the top 150 NYPD officers in the dataset which covers complaint incidents going back to 2010.
The ties are based on complaint allegations that were substantiated or in which the investigation ended inconclusively, such as when the CCRB could not get in touch with a complainant or could not prove or disprove the allegation. WNYC/Gothamist excluded allegations that the CCRB deemed to be unfounded or in which the officers’ actions were deemed to be justified. Missing from the dataset are officers who were accused of misconduct, but whom the CCRB could not identify or whose investigations were resolved through mediation.
Fox also colored distinct “communities” identified within the network by an algorithm from Gephi, a data visualization program. Unsurprisingly, many of these network communities are composed of officers who spent time in the same precincts or units. The network maps throughout this piece only name officers who have had at least one allegation against them substantiated by the CCRB. (For more on the data methodology, read EJ Fox’s in-depth explanation here.)
The resulting visualization below illustrates the extent to which controversial police conduct is often a group behavior. The following is an explorable map of the network of officers tied to complaint incidents since 2010, filtered down to only include officers with at least 99 ties to other officers. (Note: WNYC/Gothamist filtered the data down to show only actors with large numbers of ties for ease of visualization. Because of the filtering, you will not see all of the officer nodes connected to relatively large, visible officer nodes. But you will see connections to other relatively large nodes in the network.)
(The interactive graphic shows a filtered version of the network of civilian complaints against NYPD officers over the last ten years. It only includes the small subset of officers in the data who each have at least 99 ties and the names of those officers who have had a substantiated allegation of misconduct. Network visualization by EJ Fox.)
“Police misconduct can, of course, happen from a solo problem officer,” said Andrew Papachristos, a sociologist at Northwestern University, who has co-authored numerous studies on networks of police misconduct. “But the finding of these clusters of misconduct, these clusters of complaints, suggests that there's something that's actually happening in the formal and informal networks of cops.”
He added that the people an officer is connected to can affect how they "feel, think, and do"—with bad or even violent behavior being an outcome that networks can help explain.
The NYPD did not respond to requests for comment.
In a statement, CCRB Chair Fred Davie said his agency can learn from network analyses like this, particularly as it studies how to investigate racial profiling and other forms of bias-based policing.
“The NYPD must take civilian complaints of misconduct seriously because New Yorkers who experience police misconduct deserve accountability,” said Davie. “This analysis is useful for exploring the ways in which complaints of misconduct among a small group of officers uncover larger, systemic issues within the NYPD that have remained unaddressed."
The Network Outliers
The network dataset shows considerable variation on the force. Most officers, who were named in complaint investigations with any colleagues at all since 2010, were only connected to five colleagues or less. On the other hand, the top 9.35% of officers were connected to twenty or more colleagues, and frequently these officers had numerous repeat ties with a subset of those colleagues.
Many of these highly-connected officers are at the center of distinct communities in the network and have controversial histories. Several were disciplined or pushed off the force by the NYPD while a handful were convicted of criminal conduct.
One big officer node in the middle of a relatively-isolated light blue cluster at the bottom of the explorable, filtered network map represents a detective who recently retired after he was seen on camera using an alleged banned chokehold during an arrest in Inwood.
Two hops away from him in the same network neighborhood stands another large node representing an officer who was caught in an NYPD sting transporting cocaine from the Bronx to Manhattan for an undercover officer posing as a drug dealer in 2018. That officer had 254 network ties with 29 unique officers in the period Gothamist/WNYC analyzed. (Note: Most of these ties are not visible in the graphic above because the explorable network map only includes the small subset of officers in the data who each have at least 99 ties).
Pantaleo, the officer who fatally choked Eric Garner, also pops up in the filtered network at the center of a small light yellow cluster on the far right. His cluster is tied to a much larger floating clump of officers mostly from Staten Island.
Historically, Papachristos notes, many police corruption and misconduct scandals in big cities across the country have centered on teams of officers, rather than solitary bad actors.
In Baltimore, an elite gun unit carried out a series of spectacular robberies of drug dealers until they were arrested by the FBI in 2017. In 1990s Los Angeles, anti-gang officers stole drugs and framed residents, resulting in numerous overturned convictions.
“What happens in those cases is you cover for each other, you plan with each other,” Papachristos said. Because of these group patterns, the sociologist asserts, it is fair for authorities to use network mapping to identify potential collective deviance. “You can get a glimpse of, potentially, folks that are involved in those sorts of crews of cops by doing these sorts of analysis,” he said.
The officer ranked highest in the network—based on his unique complaint appearances with some colleagues and repeat appearances with others—is Robert Martinez, a decorated sergeant who worked for nearly two decades in tough assignments across Brooklyn. Martinez had more than six hundred network ties based on his co-appearances in complaint investigations since 2010 with 63 other officers.
In 2006, he and another officer in an aggressive, plainclothes unit called “Anti-Crime” were named “Cops of the Month” for pulling over a man driving through Brooklyn with a van carrying over six hundred pounds of marijuana. In 2014, Martinez won the department’s Police Combat Cross, its second highest medal, which is given for heroic acts during struggles with armed adversaries. In 2019, the sergeant received more adulation for leading a midnight team of Anti-Crime officers, who arrested an alleged gang member in East New York with two handguns after responding to reports of shots being fired.
But with all of these accolades and high-profile arrests came complaints from civilians against Martinez and the plainclothes officers working with him.
In 2007, a 23-year-old Black resident in East Flatbush accused Martinez and a colleague of getting out of an unmarked car and pushing him up against a wall, as part of a search outside his house. In one interview with a CCRB investigator, the young man claimed that, upon seeing the officers, he tried to walk away and call his mother for help. That’s when, the young man alleged, Martinez placed his elbow on his throat “trying to wrestle” him.
Police didn’t make an arrest and the CCRB substantiated a force charge against Martinez. Yet, the NYPD meted out no discipline, according to CCRB records.
Six years later, while serving as a sergeant in East New York’s 75th precinct, another Black man accused Martinez of escalating a street search.
In a civilian complaint, 33-year-old Desmon Goins claimed he had been fending off two other plainclothes officers who were attempting to search his pockets without explaining themselves, when Martinez stepped in and punched him in the face.
The two officers then helped Martinez slam him to Goins to the ground, according to a lawsuit he later filed. As Goins lay on the pavement outside his house, the suit alleges, Martinez pressed his boot down onto his face, busting his lip.
After the scuffle, police charged Goins with resisting arrest, but prosecutors dropped the charges and Goins filed a civilian complaint. After an investigation the NYPD docked Martinez ten vacation days.
In response to Goins’ subsequent lawsuit, the sergeant and other officers at the scene denied Goins’ brutality claims, according to court records. The city later settled with Goins for more than $50,000
In the years that followed, Martinez continued to generate substantiated allegations for improper stops, searches, and using foul language resulting in minor, or no discipline. In the meantime, a dense cluster of officers tied to him through civilian complaints gradually formed.
(Above, year-by-year cumulative growth of the complaint network associated with the 75th precinct; red nodes are officers, blue nodes are incidents. For more on this distinct way of mapping, read EJ Fox’s in-depth explanation here.)
Martinez did not respond to repeated requests for comment submitted to union representatives and the attorneys who represented him in Goins’ lawsuit. CCRB records indicate he left the force around the end of 2019.
Big Nodes: A Measure of Influence or Activity?
Many of the biggest nodes in the network are officers like Martinez who worked for years in high-crime areas and in specialized units like narcotics or the Anti-Crime squads. NYPD Commissioner Dermot Shea phased out Anti-Crime last summer precisely because of the teams’ controversial record.
Joseph Giacalone, a John Jay College adjunct professor and a former NYPD detective sergeant, points out that department veterans, especially those who stay on the street like sergeants, accumulate numerous opportunities for civilian contact and subsequent complaints. He also pointed out that officers in combative and team-oriented assignments are more likely to generate complaint ties.
“Any situation where there are tons of arrests, you’re going to have lots of complaints,” Giacalone said.
But even accounting for variables like unit assignments, researchers have found that key officers can exert negative influence, spreading misconduct across their networks. Examining force complaint data in Chicago, Papachristos and several other researchers found that officers working with a higher proportion of colleagues with use of force histories were more likely to be involved in future use-of-force complaints themselves.
WNYC/Gothamist was unable to do a similar analysis in New York City to assess potential peer influence because precise historical data on officers’ entry and exit from NYPD units over their careers is not currently available. (WNYC/Gothamist asked the NYPD for this data and also has an outstanding public records request seeking it.)
But several residents suggested to WNYC/Gothamist that their experiences support the contention that officer misconduct can be viral.
On an April evening last year, Adegoke Atunbi said he was attempting to record a police encounter he had stumbled upon on a street in East New York when an officer suddenly pushed away his camera and began assaulting him. Video shows an officer in a blue Under Armour shirt telling Atunbi to back up before suddenly jerking his arm towards the camera.
“He just lunged at me, he grabbed me, punched me in the face, and threw me to the ground, and a bunch of officers just tackled me,” the medical supply company owner claimed.
The officer’s badge number in the video matches that of Adnan Radoncic, a sergeant in the 75th Precinct, as The Brooklyn Paper first reported.
Radnonic has been the subject of more than a dozen civilian complaints and lawsuits over his career, and is ranked as the fifth biggest node in the network map. In the period WNYC/Gothamist analyzed since 2010, the sergeant had accumulated more than 400 network ties spread out between 68 colleagues.
Atunbi asserted that Radoncic was a catalyst for a group assault on the street that day.
“As soon as he grabbed me, all the officers was hands on,” he said. “It’s like they just followed his lead.”
While he was on the ground, Atunbi said he felt a group of around ten officers on top of him. He remembers being punched in the face and kicked in the head.
“You would have thought I had a rocket launcher in my hand, the way they tackled me to the ground,” he continued. “All I had was my cell phone.”
Atunbi is now suing the NYPD. After the encounter with police, he said, he had to have hip surgery and is still going to physical therapy to recover from his injuries.
“I can’t take two steps without holding onto something due to this incident,” he said, noting that he has to use a roller. “I can’t play with my kids. It has changed my life forever.”
Radoncic, who has received numerous commendations for “meritorious” and “excellent” police duty and made more than ninety felony arrests over his career, did not respond to requests for comment submitted to the Sergeants Benevolent Association.
Policy Implications
M.K. Kaishian, Senior Policy Counsel for the Brooklyn Defender Services, notes that the NYPD itself has the data necessary to identify and remove officers who could be spreading misconduct across the force.
“And yet what happens is the opposite, these clusters of misconduct and abuse are actually protected and further and further entrenched into the NYPD,” she said. “And that’s because they really are emblematic of the NYPD’s culture. They’re not an aberration.”
Even if certain officers form clusters simply because of the nature of their assignments, that, too, should be worrying, Kaishian argued. “That is not a form of policing that we should be tolerant of,” she said. “If they are making so many arrests, and therefore have so many victims, that’s still a problem.”
Giacalone, the adjunct professor and former detective sergeant, said that repeated complaints are certainly worth looking into, especially after investigators detect potential criminal conduct. But, he cautioned, many New York City residents in communities suffering from high crime don’t want to get rid of the specialized units wholesale, even if their active approach generates complaint ties.
He pointed to the success of Eric Adams, a former NYPD captain who is currently the leading candidate in the mayoral race. Adams has campaigned on bringing back the controversial plainclothes Anti-Crime units to address gun violence in non-white, working class communities, which are his base.
“That’s the irony of the whole thing,” Giacalone said. “You had the defund the police, anti-cop movement, and they’re actually going to … wait for this… elect the ex-cop mayor.”
Papachristos, the Northwestern sociologist, contended that firing and disciplining problem officers is an important, but limited step towards accountability.
“Unless we change and address the issues that create these networks in the first place, we're just going to kind of be repeating these sorts of patterns,” he said.
