As activists across the United States have united behind the slogan “defund the police,” with some using that to mean shifting a share of funding to social services while others call for the elimination of policing altogether, a new study points to the danger of an abrupt or haphazard withdrawal of policing.
In a working paper released this month by Harvard economist Roland Freyer and researcher Tanaya Devi, the pair found that in five major cities where deaths of African-Americans in police-involved incidents went viral and state and federal authorities subsequently placed police departments under investigation, homicides and other forms of crime surged—with almost 900 more people murdered and 34,000 felonies committed over two years in those cities “than would have been expected.”
In an e-mail interview, Freyer stressed that he “would not call this a ‘Ferguson Effect,” a term used by then St. Louis Police Chief Sam Dotson as arrest rates plummeted and crime rates shot up in his city and across the region following the killing of Michael Brown and subsequent protests and which was later amplified by then FBI Director James Comey, who suggested that the era of social media videos of police abuse is breaking down community trust and leading to a surge of violent crime.
What Freyer and Devi found is that in five cities—Baltimore, Chicago, Cincinnati, Riverside, and Ferguson—that had both viral incidents of police-involved deaths of African-Americans and subsequent state and federal investigations, police dramatically withdrew even as 911 calls stayed constant, costing hundreds of people their lives.
To reach this conclusion, Freyer and Devi analyzed what are called “Pattern-or-Practice” investigations of local crime and homicide rates. These are conducted by the Department of Justice’s Civil Rights Division—created as part of the 1994 crime bill spearheaded by then Sen. Joe Biden and signed by President Bill Clinton that also included funding for 100,000 new police officers and billions for new prisons—to “bring to light any persistent patterns of misconduct within a given police department.”
And indeed, these investigations have uncovered serious issues. For instance, after the killing of Michael Brown in Ferguson, resulting protests moved the DOJ to investigate that police department—exposing a wide range of unconstitutional practices.
Freyer and Devi found that homicide rates and total felony crimes actually dropped in most cities placed under these investigations, showing that outside probes of local policing can not only help identify abuse but can also help improve practices in a way that lowers overall crime.
But in the five cities named above, the researchers found a “marked increase in both homicide and total crime” following the investigations. What distinguished those five was a viral incident of deadly force that preceded those investigations. For instance, Chicago’s investigation began following the shooting of Laquan McDonald, a case that rose to national attention.
The number of deaths in the cities affected was sizable. The researchers estimate that the “casual effect of the investigations in these five cities… resulted in 893 more homicides than would have been expected with no investigation and more than 33,472 additional felony crimes.”
One possible culprit the study names for the increase in crimes and deaths is that “there is significant evidence that police activity changed considerably after pattern-or-practice investigations were announced post viral incidents of deadly use of force.”
While the police departments in those cities effectively chose to withdraw in response to those pressures, and weren’t disbanded or defunded, they all saw a dramatic decline in proactive policing, something I reported on back in 2018 while covering a series of events organized by activists in Baltimore to encourage peace and nonviolence amidst a homicide wave. There were 39,654 Baltimore arrests in 2014, the year before Gray was killed, compared to 25,820 arrests in 2016, a 35 percent decrease. Meantime, homicides went from 211 to 318, a 50 percent increase.
Baltimore’s police were not defunded and the scope of their lawful authority was mostly unchanged, outside of some reforms to unconstitutional practices. However, police themselves started to pull back from routine policing. “In all candor, officers are not as aggressive as they once were, pre-2015. It’s just that fact,” acting Police Commissioner Gary Tuggle said in 2018. One example of this is field interviews, where a police officer approaches someone for questioning. Between 2014 and 2017, these dropped by as much as 70 percent.
One retired Baltimore lieutenant told USA Today that no one ordered police to be less proactive but police began to fear the possible consequences of policing. "We didn't have to tell them," he said. "We just said these are the facts, this is the situation, and if you want to risk your career, have at it."
The Rev. Kinji Scott, a community activist who ran a local Baptist ministry called My Father’s House, told me at the time that he worried about the impact of a reduction in proactive policing. “We saw the police department arrest less during a period of high crime,” he said. “So what happened is you have a community of emboldened criminals.” Scott was speaking from a deeply personal place—he lost a cousin to homicide in Chicago and his brother was murdered in St. Louis.
In Chicago, Freyer and Devi found an even sharper reduction in policing following the feds’ investigation after the killing of Laquan McDonald. They note that in the “month before the investigation there were 39,074 police-civilian contacts recorded. In the month after the investigation there were 4,371 police-civilians interactions recorded – an 89% decrease.” They estimate that about 80 percent of that decrease in stops was due to the investigation, while “98% of the increase in homicides can be accounted for by reductions in police-civilian interactions in Chicago.” On June 8, Chicago had its deadliest day in 60 years, as it suffered 18 murders in 24 hours.
The Fryer-Devi study poses a dilemma, then, for citizens and public officials who want to investigate and prohibit police abuses but also protect the public’s safety.
For his part, Freyer stresses that investigating the police to stop abuses is important but that also the need to care in doing that. “In general, these investigations are a good idea—they lead to a slight reduction in homicides and total crime,” he told me. “What we have now, in intense environments with social unrest, is not working. My suggestion is to find a way to do reform with police, not to police.”
In this view, winning over police support for reforms is essential, so that they continue to engage in the kind of policing necessary to stop violent crime.
As Fryer and Devi write at the end of their study, “If the price of policing increases, officers are rational to retreat. And, retreating disproportionately costs black lives.”