For nearly a year, Richard Rosenfeld’s research on crime trends has been used to debunk the existence of a “Ferguson effect”, a suggested link between protests over police killings of black Americans and an increase in crime and murder. Now, the St Louis criminologist says, a deeper analysis of the increase in homicides in 2015 has convinced him that “some version” of the Ferguson effect may be real.
Looking at data from 56 large cities across the country, Rosenfeld found a 17% increase in homicide in 2015. Much of that increase came from only 10 cities, which saw an average 33% increase in homicide.
“These aren’t flukes or blips, this is a real increase,” he said. “It was worrisome. We need to figure out why it happened.”
All 10 cities that saw sudden increases in homicide had large African American populations, he said. While it’s not clear what drove the increases, he said, he believes there is some connection between high-profile protests over police killings of unarmed black men, a further breakdown in black citizens’ trust of the police, and an increase in community violence.
“The only explanation that gets the timing right is a version of the Ferguson effect,” Rosenfeld said. Now, he said, that’s his “leading hypothesis”.
Other experts have argued that it’s still hard to know whether 2015’s increase in murders was significant, much less what might have caused the trend. The liberal Brennan Center found that increases in homicide last year were localized in only a few cities, and that “community conditions” were likely to blame, rather than “a national pandemic”.
Even if the increase in homicide is significant, there are many competing theories for what may be responsible. The Brennan Center pointed to economic deterioration of struggling neighborhoods. Columnist Shaun King argued last month that the increase in violence in two cities seemed to be caused by police officers “refusing to fully do their jobs”. Local police officials have blamed court system failures, gang dynamics and the proliferation of illegal guns.
Rosenfeld’s new analysis of homicide trends, which was was funded by the Department of Justice, is currently being reviewed by department officials and has not yet been released to the public. A justice department spokeswoman said the paper is expected to be released in July.
The question of whether there is any link between protests over police mistreatment of black Americans and an increase in violence in some black neighborhoods has been a political flashpoint for the past year. Conservative writer Heather Mac Donald warned in May 2015 that protests over police behavior would only backfire on black citizens.
“Unless the demonization of law enforcement ends, the liberating gains in urban safety over the past 20 years will be lost,” she wrote. Her op-ed, titled The New Nationwide Crime Wave, sparked a months-long debate.
The Obama administration repeatedly denied that there is any evidence of a “Ferguson effect”, while FBI director James Comey reiterated his suggestion that violent crime was increasing because of “a chill wind blowing through American law enforcement over the last year.” Protesters said the conservative focus on the Ferguson effect is an attempt to undermine the movement to reform American policing.
Rosenfeld said that the version of the Ferguson effect he now found plausible was very different from the one Mac Donald had described.
“She thinks the solution is to stop criticizing the police; I think the criticism is understandable, rooted in a history of grievance, and serves as a reminder that the police must serve and protect our most vulnerable communities.”
If a breakdown of trust between police and community is leading a spike in murders, he wrote in an email, the solution required two things: better community policing in communities of color, and “more effective response to serious violent crime,” focused on redoubled efforts to solve homicides and other acts of violence.
Comey reignited the debate on Wednesday, telling reporters that the continued increase in violence was a serious problem that national media outlets were choosing to ignore. He said that private conversations with police officials across the country convinced him that “marginal pullbacks by lots and lots of police officers” afraid of being the subject of the next viral video of police misconduct might be contributing to the increase.
“The people dying are almost entirely black and Latino men,” he said. “It’s a complicated, hard issue, but the stakes couldn’t be higher. A whole lot of people are dying. I don’t want to drive around it.”
The White House clashed with Comey last year over his previous comments on policing and crime increases, and the administration has repeatedly pushed back against the idea of a “Ferguson effect”. Obama himself cautioned against trying to“cherry-pick” crime data last year, and Attorney General Loretta Lynch said that while the idea of the Ferguson effect had been bolstered by anecdotes, “there’s no data to support it”.
Chicago, Obama’s hometown, has seen more than 1,000 shooting incidents so far this year, compared with about 600 incidents during the same period last year. Murders in Chicago are up 56%, with 70 more people murdered so far this year than last year. Among those injured in Chicago so far this year was Zariel Trotter, a 13-year-old advocate against gun violence. Lee McCullum Jr, a former South Side prom king featured in a CNN documentary, was fatally shot on Thursday.
“The numbers are not only going up, they’re continuing to go up faster than they were going up last year,” Comey said of the uptick, according to Politico. Comey told reporters he would not call the trend he was seeing the Ferguson effect, the New York Times reported.
Some protesters and law enforcement leaders criticized Comey for advancing a theory without national data to back it up.
“Director Comey’s recent comments about a ‘viral video effect’ are unfounded, and frankly, damaging to the efforts of law enforcement,” Ronal Serpas, the chairman of Law Enforcement Leaders to Reduce Crime and Incarceration, said in a statement Thursday.
Serpas cited a series of influential reports from the liberal Brennan Center that found no change in overall crime in 2015 in the nation’s 30 largest cities, and only a slight increase in violent crime.
The Brennan Center analysis did find that the murder rate had increased 13.2% in the nation’s 30 largest cities, but it downplayed this finding. “While this suggests cause for concern in some cities, murder rates vary widely from year to year, and there is little evidence of a national coming wave in violent crime,” the report noted.
The three cities that had seen the biggest increases in murder “all seem to have falling populations, higher poverty rates, and higher unemployment than the national average,” the Brennan Center report concluded. “Economic deterioration of these cities could be a contributor to murder increases.”
Rosenfeld, a professor of criminology at the University of Missouri St Louis and the chair of a National Academy of Sciences roundtable on crime trends, said the Brennan Center’s focus on the economic roots of violence was not enough to explain “why homicide increased as much as it did in these cities in a one-year period”.
“The conclusion one draws from the Brennan Center’s report is, ‘Not much changed,’ and that is simply not true. In the case of homicide, a lot did change, in a very short period of time,” he said.
While “economic disadvantage is an extraordinarily important predictor of the level of homicide in cities,” he said, “there’s no evidence of a one year substantial economic decline in those cities. There have to be other factors involved.”
The idea of a “Ferguson effect” was coined in 2014 by St Louis police chief Samuel Dotson. The same year that Ferguson saw massive protests over the killing of 18-year-old Michael Brown, St Louis saw a 32.5% increase in homicides. “The criminal element is feeling empowered by the environment,” St Louis’s police chief argued, blaming the increase in crime on what he called “the Ferguson effect”, and arguing that the police department needed to hire 180 more officers.
That claim was picked up in May 2015 by Mac Donald, a fellow at the conservative Manhattan Institute which had published a researcher’s 1996 warning about the purported rise of “juvenile super-predators”.
Samuel Sinyangwe, a co-founder of Mapping Police Violence and Campaign Zero, called the conservative focus on the Ferguson effect “a reactionary attempt to undermine the movement”.
“It has been the attempt to put across this narrative that any criticism of the police is dangerous to society,” he said.
That kind of political rhetoric has been used against civil rights advocates in the past. Opponents of the 1964 Civil Rights Act argued that “civil rights would engender a crime wave”, Yale political scientist Vesla Weaver wrote in an article on how arguments about crime were used to attack and undermine African Americans’ fight for equal rights.
A closer look at many of the statistics Mac Donald used to bolster her thesis showed they did not provide sufficient evidence of a nationwide crime wave, criminologist Frank Zimring argued last year.
When Rosenfeld analyzed St Louis’s crime data, he found the increase in homicides there could not have been caused by a “Ferguson effect”, because the greatest increase came early in the year, months before Michael Brown’s death or the protests that followed.
But that paper only looked at the evidence for the effect in one city. With funding from the National Institute of Justice, the justice department’s research arm, Rosenfeld did a new study early this year that looked that more broadly at homicide trends in the nation’s 56 largest cities and found an overall 17% increase in homicide.
As a result of that broader national analysis he said, he has had “second thoughts” about the Ferguson effect. “My views have been altered.”
Looking at the additional homicides in large cities, he found that two-thirds of the increase was concentrated in 10 cities: Baltimore, Chicago, Houston, Milwaukee, Cleveland, Washington, Nashville, Philadelphia, Kansas City and St Louis.
Those 10 cities had somewhat higher levels of poverty than the other cities he examined. But, he said, the “key difference” was that “their African American population was substantially larger than other large cities”: an average of 41% in those 10 cities, compared with 19.9% in the others.
Separate analyses looked at two of these cities in 2015 and early 2016. A FiveThirtyEight assessment of Chicago crime data concluded that the city’s increase in gun violence was statistically significant, that the spike dated back to the release of the video of the police shooting of 17-year-old Laquan McDonald, and that it was closely correlated with a drop in police arrests. Researchers in Baltimore found a similar correlation between a drop in arrests and an increase in violence in the wake of protests over Freddie Gray’s death, and concluded that while the Ferguson effect played no role in Baltimore’s rising violence, a “Freddie Gray effect” may have been a significant factor.
Violence has many complex causes, and decades of exhaustive research has shed only partial light. Even the dramatic drop in violence and crime since the early 1990s–the most basic fact about crime in America–is not fully understood. In trying to understand 2015’s murder trends, Rosenfeld looked for reasons why cities that already struggled with high levels of violence might see “a precipitous and very abrupt increase”.
Rosenfeld considered two potential alternative explanations: the US heroin epidemic, and the number of former inmates returning home from prison. Neither of these explanations quite lined up with the increase in violence, he said. For instance, the country has been in the midst of a heroin epidemic since 2011. Why there would be a four to five year lag before the epidemic caused murders to spike?
“That led me to conclude, preliminarily, that something like a Ferguson effect was responsible for the increase,” he said.
What exactly that effect might be is far from clear, he said. The fierce debate over the “Ferguson effect” or Comey’s “viral video effect” has described the dynamic in several ways, including criminals being “emboldened” by protests agains the police, and “de-policing”, or police drawing back from proactive activities, in the wake of increased public scrutiny. One Chicago officer said that police were drawing back not because of public scrutiny via cell phone videos, but because of their fear that city officials would no longer protect officers who made honest mistakes while doing a difficult job.
Rosenfeld said he has only seen clear evidence of decreases in proactive police activity in Chicago and Baltimore. He said he believed “de-policing” was not a major factor in other cities–and that even in Chicago, changes in proactive police activity could only be responsible for some of the increase in shootings and violence.
One potential link between public attention to police violence and increased violent crime in the community, he said, might be if intensified community mistrust of the police make offenders think “that they can commit crime with impunity. They don’t think the community is willing to cooperate with the police and investigations or they think the community is less likely to contact the police when victimized.”
“We don’t yet have the data to understand the mechanism for the Ferguson effect,” he said.
Advocates for police reform and community violence prevention greeted the hypothesis with some skepticism. Dante Barry, executive director of the Million Hoodies Movement for Justice, said the idea that increased political activism by black Americans could lead to increased violence was “anti-democratic”.
Mistrust between some black neighborhoods and police was not some new phenomenon, he said. “The trust factor has always been a factor, for generations.”
Sinyangwe, of Mapping Police Violence and Campaign Zero, said he could not evaluate Rosenfeld’s research before reading his full report, but noted that New York City, which saw both massive protests over the death of Eric Garner and abrief drop in proactive police activity as a protest against Mayor Bill de Blasio, has not seen dramatic increases in crime or murder.
“The reality is we really don’t know what leads to increase in homicide,” he said. “You have to go into communities and ask them what is going on.”
Phillip Atiba Goff, a leading researcher on racial bias in policing and the president of the Center for Policing Equity, said in an April interview that one way of interpreting the Ferguson effect is “on its face, offensive”, but that there is clear research evidence linking perceptions of police legitimacy to how willing people are to break the law.
“If you believe not having police doing proactive stops in neighborhoods leads to immediate upticks in violent crime, that suggests that the people who live in that neighborhood are just waiting to commit acts of violence until they’re not being watched by the hall monitors that wear badges and guns,” he said. The suggestion that some Americans “can’t control their base instincts” without someone with a badge surveilling them is the kind of logic that led to mass incarceration and the war on drugs, he said.
“A far more reasonable hypothesis is that the decay in police legitimacy is harming both police morale and community morale,” he said. “When you don’t believe police are legitimate, you are much more likely to be engage in illegal behaviors and be uncooperative with law enforcement.”
Mac Donald, whose op-ed launched the Ferguson effect debate, said in April that increased homicides in cities with large numbers of black residents “is exactly what you would predict from the Ferguson effect. That’s where the Black Lives Matter message that police are racist is going to have the most effect, and it’s where cops are getting the most pushback from proactive policing–and where policing is most necessary.”
The Obama administration has rejected Ferguson effect because “it goes against a broader agenda of saying the real solution is social services, fighting racism,” she said.
“People are very opposed to acknowledging the connection between policing and crime,” she said. “You see the Brennan Center trying to tie itself into knots denying its own data.”
In response to criticism of the Brennan Center report, Inimai Chettiar, who oversaw the production of the report, said it “looked specifically at what’s happening to crime and not the cause. We didn’t look into whether or not there’s a Ferguson effect.”
“There have been a lot of people saying there was a crime wave and there was a murder wave nationally, and what the report’s results show is that there was not a national murder wave, there is not a national crime wave. Crime is still the same, however there are pockets that are experiencing higher murders, and those are serious things that we need to address.”
Rosenfeld blamed the FBI’s extremely slow crime data system for fueling a “largely rhetorical–largely evidence-free debate”. The FBI’s full national crime numbers for last year will not be released until September, more than a year after the debate over the year’s crime trends began.
In the absence of official national statistics, his research was based on police department data collected by the Washington Post’s Wonkblog, he said.
“That’s simply unacceptable, that’s unnecessary, and the FBI really needs to get its act together,” he said.