“Predictive Policing” and Racial Profiling
While technology used in policing has improved, it hasn’t progressed, says Khalil Gibran Muhammad, if racial biases are built into those new technologies. This excerpt from his book, The Condemnation of Blackness: Race, Crime, and the Making of Modern Urban America, shows that for the reform called for by the current protests against systemic racism and racially-biased policing to be fulfilled, the police — especially those at the top — will need to change their pre-programmed views on race and the way they see the Black citizens they are supposed to “serve and protect.”
Law enforcement, especially, has doubled down on crime statistics in what is now the era of big data, artificial intelligence, and predictive analytics. Old ideas, yet again, have been programmed into the latest technology.
At the 2015 New York Times Cities for Tomorrow conference on the newest advances in technology and data analytics for everything from urban-based environmental sustainability to crime control, then Police Commissioner William J. Bratton spoke about the New York Police Department’s latest crime-fighting tool. I sat in the audience anxious to hear him speak. With a broad smile and supreme confidence, he praised the newest release of the pioneering crime mapping software known as CompStat, which had been at the heart of stop and frisk policing when it began a generation ago. He likened the latest version to the 2002 film Minority Report, starring Tom Cruise as head of a special precrime unit. Set in Washington, D.C., in the year 2054, officers gathered intelligence from a trio of precogs, humanlike beings who can predict murders and identify killers before they act. Bratton was almost giddy about the comparison; the unintended pun on the film’s title seemed to escape him.
Two years later at a 2017 Heritage Foundation summit on “Policing in America: Lessons from the Past, Opportunities for the Future,” Bratton gave more details about the architecture of CompStat 2.0. The software is based on algorithms and “advanced data mining techniques, we call ‘predictive policing,’” he said. “Effectively, it’s the CompStat of the ’90s on steroids in the 21st century.” And just like all new technology promises, it was guaranteed to be better than before. “It is discriminating, not discriminatory,” he bragged. “It is precise, not prejudiced.”
Until he retired, Bratton was known as America’s Top Cop. Starting his career as a military police officer in Vietnam and then onto Boston in the 1970s, he spent the next five decades running the biggest and most racially troubled police agencies in the country. Bratton served in six departments coast to coast, from New York to Los Angeles and back to the Big Apple. Several of these departments were subject to federal investigations for police brutality either before or after he left. Over the years, he developed a strong personal sense of history, covering the entire span of the post–civil rights era in policing.
But unlike the many critics of aggressive policing tactics, Bratton has rarely, if ever, publicly questioned the value of social science data, except when the research critiqued police racism. In his Heritage speech, he celebrated the theoretical founders of broken windows policing, the criminologist George Kelling and the political scientist James Q. Wilson, “two personal heroes of mine.” He also repudiated the Kerner Commission findings, which he said he had read in 1974 to pass the sergeant’s exam for the Boston Police Department: “They believed at the time that the causes of crime were racism, were poverty, were police practices in many instances, unemployment, demographics. They thought those were the causes. They were not. They are not. And they never have been.” Bratton’s emphatic dismissal of the Kerner report, and all the published evidence of police bias since, demonstrates just how enduring Hoffman’s original innovation with racial crime data has been.
At the summit, Bratton said that by the 1990s, the policing profession had finally figured out how to get past the flawed Kerner legacy. Police leaders started coming together at a series of executive sessions led by faculty at the Harvard Kennedy School. There, “we began to get it right,” he noted. “The cause of crime is people.” Paraphrasing Al Gore, he said “there is an inconvenient truth” that cops go where the criminals are. “Data-driven or evidence-based policing is not bias policing.” He continued: “The disparities are not a policing issue. It is about behavior. You have the crime numbers and they are self-evident.”
Future historians will have to place Bratton’s legacy in its fullest context. For now, what’s clear is that he has had an oversized influence on how racial crime data continues to shape the lives of African Americans and Latinx people in the post–civil rights era. But of course, he was not alone. Near the end of New York City Mayor Michael Bloomberg’s third term, when activists were demanding an end to stop and frisk and the New York Civil Liberties Union and Center for Constitutional Rights were suing the city, Bloomberg refused to change the policy. He consistently dug in, insisting that racial disparities in stops were not evidence of bias, but of criminality. And, at times, he ridiculed critics for not understanding how crime data works. “In that case, incidentally, I think, we disproportionately stop whites too much and minorities too little,” he said in late June 2013 on his weekly radio show. “It’s exactly the reverse of what they’re saying. I don’t know where they went to school, but they certainly didn’t take a math course, or a logic course.”
Less than two months later, the day after a federal judge ruled that stop and frisk was racially discriminatory and unconstitutional in the case of Floyd, et al. v. New York, Police Commissioner Ray Kelly went on national television to defend racial profiling. He suggested that black and brown people, subjected to stop and frisk policing, could not be innocent no matter that they had not broken the law. David Gregory, host of NBC’s Meet the Press, asked him about the 4.4 million stops of New Yorkers, over 80 percent of whom were black or Latino, and 88 percent of whom were not even subject to a summons or arrest, between 2004 and 2012. Gregory said, let’s start with the nearly nine out of ten people “not doing anything wrong.” Kelly responded, “It doesn’t mean that people are not doing anything wrong. If you look at the statute, it says reasonable suspicion that individuals may be about, are committing, or have committed a crime.” Like precogs, the police were acting on data-driven predictions based on Compstat. “There’s a preventive aspect to this. People say innocent,” he said, stammering his way to an explanation. “That’s not the appropriate word.” Their crimes: living while black and brown. “This, by the way, is the standard law enforcement practice throughout America.”
Bratton, Bloomberg, Kelly, and so many other police and elected officials continue to defend their vision of policing today based on what they claim the data says and tells them to do. Former New York Mayor Rudolph Giuliani put it this way: “When I assigned police officers with Commissioner Bratton and Commissioner Safir, we did it based on statistics. We didn’t do it based on race. If there were a lot of murders in a community, we put a lot of police officers there.” In the name of saving black people from themselves, they’ve turned policing into the most important legacy of the civil rights movement. “In our country, the first obligation of government is public safety,” Bratton told the Heritage audience. But policing has always been at the heart of civil rights activism and the fight for equal citizenship. That is the civil rights movement’s most enduring legacy and unfinished business. Even to take them at their word, by their own empirical standards, there is no research consensus on whether or how much violence dropped in cities due to policing.
[…]
The link between race and crime is as enduring and influential in the twenty-first century as it has been in the past. Violent crime rates in the nation’s biggest cities are generally understood as a reflection of the presence and behavior of the black men, women, and children who live there. The U.S. prison population is larger than at any time in the history of the penitentiary anywhere in the world. Nearly half of the more than two million Americans behind bars are African Americans, and an unprecedented number of black men will likely go to prison during the course of their lives. These grim statistics are well known and frequently cited by white and black Americans; indeed for many they define black humanity. In all manner of conversations about race — from debates about parenting to education to urban life — black crime statistics are ubiquitous. By the same token, white crime statistics are virtually invisible, except when used to dramatize the excessive criminality of African Americans. Although the statistical language of black criminality often means different things to different people, it is the glue that binds race to crime today as in the past.
