Why you should care
At least 10 organizations track U.S. police violence, and they’re increasingly partnering with the cops themselves.
For Brandon Anderson, police violence is personal. In 2007, his partner was stopped by an officer in Oklahoma City, falsely accused of stealing a car and beaten to death. Scarred but determined, Anderson now works to prevent others from suffering a similar fate. He launched an app called Raheem last month in Oakland, California, that allows users to self-report instances of police violence they see — where it happened, the officer’s name and details of the incident.
It’s just one of at least 10 platforms and organizations cropping up across America that collect and analyze data on police encounters. These initiatives have all emerged over the past five years, following the spate of killings of African-American men at the hands of the police. And many are even partnering with police departments and oversight boards in a bid to ensure justice and influence policy.
Among them is We The Protestors, an organization co-founded by Samuel Sinyangwe — who has a data science background — with colleagues DeRay Mckesson and Brittany Packnett after 18-year-old Michael Brown was fatally shot by police in Ferguson, Missouri, in 2014. When Sinyangwe first tried to find police data to understand the magnitude of the problem, he found the available information was woefully inadequate. They launched Mapping Police Violence (MPV), a project that uses obituaries from local newspapers, media coverage and police reports to track down police killings and code them onto a visually digestible map. The platform, which experts say is America’s most comprehensive database of police violence, shows that 25 percent of the 1,147 people killed in police encounters in 2017 were Black, though the community constitutes only 13 percent of the national population.
Different platforms are filling in other data gaps. The Center for Policing Equity (CPE), a think tank in Salt Lake City, publishes data on how frequently officers use force on the job and how those actions differ according to the race of the people involved — including those of the cops. The Police Data Initiative in Washington, D.C., launched in 2015, releases data from 130 law enforcement agencies across the country. Since 2015, Rayid Ghani, a data scientist and University of Chicago professor, has collected and analyzed data from the Charlotte-Mecklenburg Police Department (CMPD) in North Carolina. Often, these platforms are able to build more comprehensive data sets than the police departments themselves: During a pilot project in San Francisco in 2017, Raheem collected twice the data in three months than the city did in one year.
Reporting the police — positive or negative — should be as easy as ordering takeout.
Brandon Anderson, Raheem co-founder
And while they’ve faced some pushback — including from police unions — these initiatives have over time found several police departments, from Orlando to Oklahoma City, willing to partner with them to improve their standards. The partnerships are helping these departments regain credibility among the communities they serve. The data is also helping them better identify threats and difficult situations police officers face — and determine what works and what doesn’t. It helps that police departments are also turning to data as a crime-busting ally, so number crunching is no longer as alien to cops as it once was.
“Our goal is to reduce the bias that is interjected into the system by police,” says Chris Burbank, the former police chief of Salt Lake City who is now vice president of Strategic Partnerships for CPE. “We want to know what police are doing to cause violence.”
It hasn’t been easy. Orlando, Sinyangwe’s hometown, has one of the highest rates of police violence in the country. When We The Protestors brought this to the attention of the police chief there, his response was that Orlando has a lot of tourists. The tourists get rowdy and require officers to use force, he said. But the data shows that Orlando has a higher rate of police violence than Las Vegas or New Orleans, which also have very high rates of tourism, says Sinyangwe. The organization’s evidence-based approach worked. In 2015, the Orlando Police Department (OPD) adopted stricter use-of-force policies for their officers by requiring them to sign a directive declaring they could be prosecuted for intentionally abusing a suspect. OPD saw a 16 percent drop in use of force that year.
Data has also helped establish what Black communities instinctively knew, he says. “After the Baltimore uprising, we could show this was more of a pattern of practice because you could see that the Baltimore police had mostly killed Black men.” They then showed the same was true in Cleveland, he says.
Raheem hopes to fix another problem. America has about 18,000 police departments, each with their own ways of reporting misconduct. “The system is incredibly decentralized,” says Anderson. Self-reporting can create a more robust database. Raheem also connects you to an attorney if you are a victim of police violence to help you file a formal complaint against an officer, file a lawsuit or talk to a journalist. Raheem is already working with the Oakland Police Commission — the police oversight board — which will use the platform’s data to address concerns about policing in the city. The Alameda public defender’s office is also using Raheem’s data to better understand police conduct in the county and identify people who might be willing to testify against officers in court. “Reporting the police — positive or negative — should be as easy as ordering takeout,” Anderson says.
CPE works with 130 police departments around the country. Its National Justice Database collects crime data, demographics from the U.S. census and data on police behavior. The data helps them provide tactics for officers to better connect with and serve communities. After years in policing, Burbank believes data is the only way to make real change. “We’ve done lots of training over the years — we called it racial sensitivity training,” he says. “But it didn’t change [police] behavior.”
Not everyone in the law enforcement community supports these initiatives. “The resistance has come from the police unions,” says Sinyangwe. “They’ve been the primary force resisting any effort to implement evidence-based reforms.” But Anderson suggests that it’s often more complex than that. He says a police chief in a California city with more than 1 million residents found merit in Raheem, for example, but expressed deep concern about publicly releasing information about officer conduct before they had a chance to review it. That city is yet to adopt Raheem.
But these platforms are also showing critics how data can help officers and departments. MPV data, for instance, shows departments that do limit the use of force are actually safer and it doesn’t restrict officers from doing their job, Sinyangwe says. Campaign Zero, another We The Protestors platform, works to create buy-in from legislators. It has a tool that allows people to directly contact their representatives with information and suggestions on curbing police violence.
Meanwhile, Ghani’s work at the University of Chicago is demonstrating how data can help departments avoid violence. He found that when an officer is repeatedly dispatched to a highly emotional situation, like a domestic abuse case or a suicide attempt, they’re more likely to use unjustified force in the period following that incident. “They need counseling,” Ghani says. “They shouldn’t be dispatched to those situations again.” Today, the CMPD in North Carolina uses Ghani’s system to stay alert to scenarios where officers are more at risk of using force.
Those partnerships are critical. Data’s no use if the officer with a gun doesn’t care for it. Nationally, the number of police killings continues to rise. But with more and more police departments now embracing these tools, data has a chance of helping resolve a key source of civic tension in modern America. And when things do change, we won’t need to rely on the government’s word for it. The data will tell us.