Several years after the extensive debates about racial profiling as part of New York City’s stop-and-frisk policy, Americans are again discussing the relationship between race, policing and crime. National news has been roiled by, on the one hand, first-person videos of unarmed black men shot by police officers, apparently without provocation, and, on the other hand, two lone shooters targeting and killing police officers in Dallas and Baton Rouge.
In response, some are forecasting a wave of protests and riots in African American communities, mirroring the events of 1968. Others predict a looming increase in violent crime, positing a “Ferguson Effect,” in which the national debate over racial justice leads to more public scrutiny of police officers, which prompts officers to step back in their policing, allowing crime to escalate.
Are they right? And if so, what can be done?
The “broken windows” theory behind increased policing is still just a theory
While intuitive and appealingly simple, it remains an open question whether expanding law enforcement unfailingly reduces criminality. If true, increasing the number of police and the scope of arrests would reduce crime and improve ordinary citizens’ lives. That’s become known as “broken windows” policing — the idea that catching small, quality-of-life violations (like illegally selling individual cigarettes or CDs) will keep a neighborhood safe from more serious crime.
Let’s look at the evidence.
Research shows that “broken windows policing” overwhelmingly targets minority communities. Some recent research suggests that race may not motivate police shootings. But scholarship shows overwhelmingly that U.S. police are more likely to arrest and use force on black and Latino civilians than on whites.
What’s more, governments send police to areas where they expect more crime, which are often minority neighborhoods. Police stop more people and make more arrests in the neighborhoods they actively patrol — and so, by their presence, they push up crime statistics. That feedback loop leads to an ongoing focus on minority neighborhoods.
Even law enforcement doesn’t seem to know what works. For instance, FBI Director James B. Comey suggested that a declining police presence would lead to increased crime but later acknowledged that he “needed better data.”
We found a natural experiment that reveals some surprising insights
How can we get that data? It’s not possible to do an ordinary controlled experiment in which we take two identical neighborhoods and use one set of tactics in one of them and a different set in the other. So we looked at a natural experiment that occurred in New York City in late 2014 and early 2015, when the New York Police Department, in a fight with Mayor Bill de Blasio, held a work slowdown and reduced their policing. If the theory is correct — if more policing reduces crime — then less policing should lead to increased crime.
Here’s the background. As many will recall, in November 2014, a St. Louis grand jury declined to indict Darren Wilson, the officer who had shot and killed unarmed teenager Michael Brown in Ferguson, Mo. In response to the decision not to indict, organizers held massive protests in major cities. In New York, thousands of protesters marched across the Brooklyn Bridge, while others blocked portions of the West Side Highway, as well as the Lincoln and Holland tunnels.
In the midst of these protests, an unstable young man named Ismaaiyl Brinsley fatally shot two young NYPD police officers, Wenjian Liu and Rafael Ramos, before escaping to the subway and shooting himself. For a variety of reasons, the city’s police unions blamed the shooting on de Blasio’s comments about the deaths of Michael Brown and of a New York man, Eric Garner, who also died at the hands of police.
Here’s where we get to the experiment. NYPD officers are legally prohibited from striking. So they unofficially coordinated a long-term slowdown in policing. Officers were explicitly instructed to respond to calls only in pairs and to perform only their most crucial duties. Indeed, they were told not to leave their squad cars unless they felt compelled.
In practice, the slowdown meant that from the beginning of December 2014 through early January 2015, officers were significantly less likely to enforce minor legal violations, though they continued to investigate more serious crimes, such as homicide, rape and grand larceny.
Over this period, summonses (which include all ticketable offenses, such as parking fines as well as disorderly conduct and drinking in public) were down more than 40 percent from the same period the previous year, while minor arrests were down more than 25 percent. During the week directly after the shooting of Ramos and Liu, the slowdown was even more pronounced: Total arrests were down 66 percent from the same period a year prior, while traffic citations dropped 94 percent. A month after the slowdown began, summonses were still 40 percent lower than the previous year’s.
Here’s what our analysis found
If more policing reduces crime, then we would expect less policing should lead to more crime.
But in fact we find the opposite. Civilian complaints of major crimes — murder, rape, felony assault, burglary and grand larceny — actually declined during the slowdown.
We focus on major criminal complaints for two reasons. First, because these acts so severely impact the victims’ lives, we have no reason to suspect that the reductions in foot patrols would prevent citizens from registering complaints with NYPD by, for instance, calling 911 or their local precinct. Second, the premise behind “broken windows” theory is to prevent precisely these types of major crimes by arresting people for relatively minor offenses. Yet when summonses and arrest rates plummet, we see no increase in major criminal complaints.
Reducing police presence didn’t apparently lead to more crime — at least during this period.
We didn’t see any differences in criminal complaints between majority-white and majority-minority police precincts. In the minority neighborhoods that had been subject to higher levels of policing — by which we mean surveillance, quality-of-life arrests, and all the other aspects of broken windows policing — major crime did not increase.
Indeed, it could be argued that being subject to fewer summonses and arrests for minor offenses might actually increase the quality of life for minorities who live in disproportionately targeted communities.
Christopher M. Sullivan is an assistant professor of political science at Louisiana State University. Zachary P. O’Keeffe is a PhD student in political science at the University of Michigan.