It's become too clear to ignore: In many American cities, homicide is on the rise. The 56 largest cities saw 17 percent more homicides in 2015 than in 2014 — and in 10 of those cities, homicides were up by more than 60 percent.
For people who've been following the debate around police-community relations in America, though, this isn't a surprise. For the past year, defenders of police have been raising the specter of the "Ferguson effect": the theory that protests against police shootings and brutality have made police afraid of doing what they need to do to keep communities safe, which has led to a rise in violent crime.
The biggest proponents of the Ferguson effect are conservatives like Heather Mac Donald (whose new book on the subject is literally called The War on Cops). But even FBI Director James Comey has put forward a version of the theory (to the administration's chagrin). And while criminologists were originally skeptical, some of them — including one who's doing an analysis for the Department of Justice — are beginning to believe there's more to the theory than originally thought.
The problem is that calling it the Ferguson effect strongly implies that if crime goes up because police are retreating, that's the fault of the public for making police scared. That polarizes the conversation before it's begun. The fact of the matter is that in many cities, crime is going up and police are retreating — and in some of those cities there's a connection between the two. But police are the only ones who can fix it.
Talk of the Ferguson effect started before there was much evidence of a change in crime rates
The phrase "Ferguson effect" lumps together three different questions: what is happening to crime in US cities; why it's happening; and whose fault it is.
The confusion between the three isn't a coincidence. People started talking about a crime wave last year, when there was barely any evidence to bear out the theory, as a way of accusing Black Lives Matter and other protesters of eroding public safety.
In May 2015 — about a month after the killing of Freddie Gray in Baltimore caused several days of protest and unrest — researcher Heather Mac Donald of the Manhattan Institute published a Wall Street Journal commentary called "The New Nationwide Crime Wave."
Mac Donald used homicide and shooting data from some cities (including Baltimore) to argue that after 25 years of falling crime rates in the US, violent crime was on the rise again. The reason for this, she wrote, was that police were afraid to do what they needed to do to keep communities safe — because they were worried about what Mac Donald called "intense agitation against American police departments":
A handful of highly publicized deaths of unarmed black men, often following a resisted arrest—including Eric Garner in Staten Island, N.Y., in July 2014, Michael Brown in Ferguson, Mo., in August 2014 and Freddie Gray in Baltimore last month—have led to riots, violent protests and attacks on the police. [...]
“Any cop who uses his gun now has to worry about being indicted and losing his job and family,” a New York City officer tells me. “Everything has the potential to be recorded. A lot of cops feel that the climate for the next couple of years is going to be nonstop protests.”
To describe this phenomenon, Mac Donald used a term coined by St. Louis Police Chief Sam Dotson back in November 2014: the "Ferguson effect." In Dotson's original formulation, the Ferguson effect referred to both the strain on police officers who'd had to put in time policing protests in the area and the idea that "the criminal element is feeling empowered" thanks to officer fatigue.
This was both specific to St. Louis and mostly bogus — subsequent analyses showed that crime was up in the St. Louis area in early 2014, before Michael Brown was shot. In Mac Donald's hands, though, it became something more universal and more credible: the idea that the fear of something like the Brown shooting happening in their areas was making officers unwilling to take necessary actions to keep their communities safe, and criminals were flourishing in the vacuum.
That's also the way FBI Director James Comey has put it on more than one occasion (while Comey resists the term "Ferguson effect" itself). He describes it as a "chill wind blowing through American law enforcement": a change in police officer behavior, driven by fear of being caught on video doing something controversial.
Homicides really are on the rise in many US cities
Mac Donald's article was readily embraced by pro–law enforcement conservatives. But it wasn't given much credence by criminologists, who accused MacDonald of cherry-picking her data and of being way too quick to identify a few data points as a trend. On the whole, they argued, crime was still down, so looking at a few specific cities to prove a hypothesis was essentially singling out statistical noise.
But over the past year, evidence has begun to mount that there really is something going on.
Preliminary FBI estimates indicate that violent crime was up 1.7 percent nationwide from the first six months of 2014 to the same period in 2015. That's not unprecedented during the current era of falling crime rates — something similar happened three years earlier — but it's moderately worrisome.
A report by the Major Cities Chiefs Association, meanwhile, collected data from 63 agencies and found that homicides were up 9 percent from the first quarter of 2015 to the first quarter of 2016 — with non-fatal shootings up 21 percent during the same period.
All of this could still turn out to be statistical noise. (It's not a coincidence that the conversation around whether a Ferguson effect–style crime wave is real is happening at the same time as a conversation about just how bad crime data is at the national level.) Where there really does seem to be something going on that's too significant and too sudden to be mere coincidence, though, is when looking at certain individual cities.
Richard Rosenfeld of the University of Missouri St. Louis — a respected criminologist who was an early critic of Mac Donald and Comey's versions of the Ferguson effect theory — has conducted an analysis of his own for the Department of Justice. As Lois Beckett writes in the Guardian, he's been surprised by what he's found.
In the 56 biggest cities in America, homicides went up 17 percent in 2016. Furthermore, most of that rise was due to about 10 cities. Some of those cities had had high-profile police shootings and protests (Baltimore, Chicago, St. Louis); some had not. But all of them had much larger black populations than other cities. The 10 cities most responsible for the homicide jump had a population that was on average 41 percent black; the other 46 cities were, on average, less than 20 percent black.
The reasons for rising violence vary city by city — but in some cities, they're definitely tied to changes in policing
On the first question, at least — the question of what is happening — the Ferguson effect theory is narrowly correct, at least in some cities. That brings up the second question: why it's happening.
To Mac Donald — and to Comey, even though he's disavowed the term "Ferguson effect" itself — the answer is simple. Homicides are going up because police are less willing to do their jobs effectively.
Rosenfeld puts it more specifically: Police officers are less likely to engage in what's called "proactive" policing. They're less willing to spend any more time in communities than is absolutely necessary to address crimes that have already happened.
Just like there's been a certain reluctance to admit homicide is rising at all among people who don't want to blame Black Lives Matter protesters for it, there's been reluctance to attribute any rise in homicides to changes in policing. The Brennan Center, for one, argues that the cities where homicides have gone up are simply more economically depressed (something Rosenfeld points out didn't start last year, and therefore can't explain a single-year jump).
It's tempting to believe that police have no effect on crime. That's what led some liberals, for example, to celebrate the NYPD "slowdown" of January 2015, in which police officers refused to make arrests for minor crimes, respond to police calls alone, or spend any more time than usual out of their cars.
But the reality is that changes in policing do affect crime rates. Indeed, "proactive" policing — in forms that have officers walking around neighborhoods and building relationships with their residents — is one of the most effective things a city can do to prevent crime. You just have to look at the correct scale: Police departments are local institutions, and they affect things on a local scale.
"Gun violence is very local," says crime analyst Jeff Asher. "And changes in gun violence patterns probably have local explanations." So he doesn't give much credence to Comey's version of the Ferguson effect theory — that the hypothetical fear of being the subject of a viral video somewhere is changing how cops around the country do their jobs. "There's little evidence in the places we can measure it," he says, "that proactivity in, say, Louisville, went down because of events in St. Louis or Baltimore."
But there is evidence that events in Baltimore — namely, the death of Freddie Gray and ensuing protests — made police in Baltimore less proactive. Asher charted shootings in Baltimore (the red line) against police proactivity, as measured by drug arrests (the blue line):
The same is true in Chicago, after video was released of the death of Laquan McDonald. An analysis that Asher did with Rob Arthur for FiveThirtyEight showed two main effects: an increase in homicides and nonfatal shootings, and a decrease in arrests for those same crimes.
Less "proactive" policing is another word for police officers not doing their jobs
The connection between community protests and a less proactive police force is actually better established than the relationship between community protests and higher crime. A 1999 study by criminologist Robert Ankony found that when police feel more alienated from, and negatively toward, members of the community, they're more likely to retreat from "proactive" policing and do only what they need to do to respond to crimes. And in some cities, like Minneapolis, police proactivity dropped in 2015 after a controversial police shooting — it just didn't affect homicide and shooting rates, which were already high.
New York's 2014 "slowdown" illustrated just how this sort of thing can happen: After the murders of officers Wenjian Liu and Rafael Ramos, reports circulated that union officials were telling officers that it wasn't safe out there and they needed to protect themselves first. "These are precautions that were taken during the 1970s when Police Officers were ambushed and executed on a regular basis," one email sent to NYPD officers (attributed to the Patrolmen's Benevolent Union, one of the police unions) read.
This sort of thing appears to confirm a belief common to critics of law enforcement: that police officers put their own safety first and community safety second.
For many officers, it appears to be correct — and it's understandable. As several former law enforcement officers have told me, it's hard to do your job if you don't get home alive at the end of the day.
The problem is that police officers aren't actually the ones in more danger from rising homicide rates. The FBI's preliminary 2015 stats show that 20 percent fewer police officers were killed on duty in the first half of 2015 than in the first half of 2014 (an unusually deadly year for officers), and the second lowest for the first six months of any of the last 12 years.
Instead, just as the blackest, poorest members of the community are the most in danger from overpolicing, they're the ones left vulnerable when law enforcement indulges in underpolicing. This isn't a new phenomenon either — it's been consistent even through the quarter-century-long drop in crime.
Even as police are getting way fewer murder cases to investigate, they're solving the same share of them: 40 percent. And the tendency to only interact with citizens when absolutely necessary, out of alienation and fear, cuts down on the relationships that (as Comey identified) are crucial in helping solve murder cases — while encouraging police to get aggressive when they do make arrests, since they have no way of knowing how dangerous someone really is.
It's revealing that the reason the discussion of rising homicide rates isn't actually centered on the effects on the community (despite some concern trolling from Mac Donald and company), but rather on the effects on police. Indeed, the Major Cities Chiefs Association's report on rising crime rates came right before National Police Week — which focuses on the dangers police face in the line of duty. But when police put their own safety first, something is actually lost.
The implications of the Ferguson effect: Either the public needs to support police officers unconditionally, or police need to find better ways of going about things
This gets at the third question that the Ferguson effect obscures: In the places where police are being less proactive, whose fault is it?
To Heather Mac Donald, the answer is clear: It is incontrovertibly the fault of critics of the police. "Unless the demonization of law enforcement ends," Mac Donald wrote in her Wall Street Journal commentary, "the liberating gains in urban safety over the past 20 years will be lost."
That's because Mac Donald simply treats it as a given that officer morale is important to good policing, and that criticism hurts officer morale. In other words, it's not just true that police officers put their physical safety first — they put their self-esteem first. And that's a good thing.
When you put it that way, it sounds tremendously insulting to cops: They can't do their jobs if they are suffering from hurt fee-fees. Indeed, when FBI Director Comey advanced his "chill wind" theory, the Fraternal Order of Police caught the less-than-flattering implication and voiced their disapproval.
It's also kind of terrifying, in its implication that cops can only keep the peace if citizens don't look too closely at how they do it: a "Rough men stand ready to do violence" ideal of what law enforcement is supposed to be.
You can't force people to support police. And despite the worries that new policies will make it harder for officers to do their jobs without "red tape," the status quo — which is arguably too good at protecting police — seems pretty stable. So in the world we live in now, there's exactly one solution to the fear of proactive policing: Police need to find ways to do their jobs that don't put them at odds with the community.
This is already happening. As Comey told University of Chicago law students in 2015, part of the change in police officers' behavior is "to be welcomed, as we continue to have important discussions about police conduct and deescalation and the use of deadly force."
One of the examples of insidious low morale Mac Donald cited in her Wall Street Journal article came from a former San Francisco deputy chief: “Officers are trying to invent techniques on the spot for taking down resistant suspects that don’t look as bad as the techniques taught in the academy." That doesn't sound like the most effective way to learn deescalation, but it doesn't exactly seem like a bad idea.
Many criminologists accept that one of the most important parts of policing effectiveness is "procedural justice," the ability of community members to see that they're being treated fairly in interactions with police. A community protest after a police shooting is a sign that the community didn't trust the department to begin with — that the procedural justice wasn't there. A retreat from proactive policing in the wake of such an incident is both a sign of the failure of procedural justice and a perpetuation of it.
If you define the Ferguson effect as a rise in homicides tied to a lack of proactive policing, then the Ferguson effect is real — in a handful of cities. If you define it as a retreat from proactive policing due to fear of public consequences, then it's real in a handful of cities and, possibly, in the heads of scattered law enforcement officials around the country.
But something that's lodged itself in the heads of officers isn't going to go away on its own. The one thing the "Ferguson effect" is not is the thing it's most commonly used as: an attempt to blame black protesters for rising homicide rates in black communities.