Stop-and-Frisk Didn't Make New York Safer

When former NYPD Commissioner Ray Kelly was asked what would happen if stop-and-frisk were curtailed, his response was characteristic of his tenure: “No question about it,” he said “violent crime will go up.” When homicides rose in Chicago, Chicagoans clamored for NYPD-style stop-and-frisk. The same premise is repeated by proponents of stop-and-frisk throughout Daniel Bergner’s illuminating Atlantic article: if you want to reduce crime, you have to be willing to suffer more aggressive policing tactics.

In reality, there’s no good reason to assume that these strategies work to reduce crime. David Greenberg has conducted the most comprehensive analysis of the relationship between the NYPD’s practice of stop-and-frisk and crime levels to date, and he finds “no evidence that misdemeanor arrests reduced levels of homicide, robbery, or aggravated assaults.”

No one thinks a police officer with a reasonable suspicion that a suspect has a gun should be barred from frisking the suspect, but that is not what stop-and-frisk has come to mean. The now-abandoned practice of requiring officers (often fresh out of the academy) to meet performance goals for citations and arrests seems wrong on several levels, but the most fundamental one is that it doesn’t reduce crime. A close second is the increased costs to families and communities. As Bruce Western, Amanda Geller, Christopher Wildeman, and many others have described, the collateral damage from broad criminalization is far-reaching, and concentrated on the populations that can least afford them.

So why are so many so enamored of these dubious tactics?

Stop-and-frisk proponents, like former New York City Mayor Michael Bloomberg, former NYPD Commissioner Raymond Kelly, and the criminologist Alan Zimring believe that the method is effective because of what they have seen in New York. Crime dropped precipitously in the 1990s, they say, and the reason for it is the distinctive way the NYPD practiced policing.

This is an argument that—in both style and substance—runs counter to modern methods of empirical research. The NYPD did not implement large-scale trials to isolate the effects of stop-and-frisk. nor did they look to any form of multivariate analysis examining the largest cities in the country. The argument is entirely based on people eyeballing statistics and telling stories.

So what about those stories? The simple version is that the NYPD adopted order-maintenance policing, including stop-and-frisk, and crime went down. But the increase in frisks and arrests didn’t predate the drop in crime; it came after the drop in crime. If we widen our perspective, we see that the crime drop in New York City, particularly for property offense, began long before 1990.

What about homicide? After all, citing that link was the most effective claim Bloomberg and Kelly used in support of stop-and-frisk. Well, here’s a piece of data you won’t hear any of the proponents tout: if you look at the 25 largest cities in the United States, only five had significantly higher homicide rates in 2010 than in 1960. Big cities—including New York City—are pretty much back where they started before the massive late-20th century crime-wave. Some of these cities did not see the stunning improvements in homicide rates that New York experienced in the 1990s, but that’s because they didn’t see dramatic increases in homicide rates in the preceding decades.

In cities where homicide rates did increase between 1960 and 1990, many display similar patterns to New York City. LA, Houston, San Antonio, San Diego, Dallas, Jacksonville, and Charlotte saw homicide rates move nearly parallel to New York City’s, although they started from, and returned to, different prior levels. Other cities, however, have homicide rates that are even lower than they were in 1960.  And still others started with higher rates and ended with lower ones.  You only get the pattern proponents describe by looking arbitrarily at particular cities and decades.

When you look at all the data and all the research, the puzzle is not so much why crime dropped in New York City, but what caused the great American crime wave in the first place. There are plenty of prominent theories. I’ll mention just two.

Lead. One possible factor is the widespread distribution of lead through motor vehicle exhaust, paint, and lead pipes —a distribution that was disproportionately concentrated in low-income inner-city neighborhoods. Children in New York have long been at particular risk compared with the rest of the country. The majority of the city's housing units were built before 1960, when the city phased out residential lead paint, and low-income families are the most likely to live in homes where the paint is deteriorating. Over the past two decades, health officials have made concerted efforts to address this problem. But the well-reported fact that inner-city families were exposed, and that two generations of low-income urban American suffered impaired cognitive development as a result, is tragic beyond measure. The close correlation between lead levels and crime rates—not just in New York, but in countries around the world—is brutal in its scope and implications.

It’s true that lead levels can’t definitively explain all the fluctuations in crime levels of the past 60 years. But think of it this way: there’s far better evidence that lead exposure caused an increase in crime than that stop-and-frisk caused a decrease. When it comes to lead exposure, the evidence supports a strong link: We know it impairs executive function by reducing brain growth in ways that are now widely studied, and research has linked individual lead exposure with increased criminal justice involvement. In contrast, stop-and-frisk has been found to have little or no effect on crime. Rather, the form of policing practiced in the last decade of Kelly’s tenure was shown was shown to produce a host of negative effects related to trust and cooperation with law enforcement, as well as a reluctance to report crimes born from sourness toward the police.

Abandonment. The historical evidence ignored by stop-and-frisk proponents also describes widespread changes in police practices across the country. In the years leading up to and during the crime wave, police effectively abandoned disadvantaged communities, prompting critical commentary like Public Enemy’s chart-topping rap song “911’s a Joke.” Read the first-hand accounts by police from that period, or the descriptions in historical accounts by Bill Stuntz and Randall Kennedy, and you will come away with a detailed account of how we failed to protect law-abiding citizens in historically disadvantaged communities throughout the ’60s, ’70s, and ’80s.

Abandonment not only contributed to the rise in crime in New York and many other cities across the country, but as Randall Kennedy noted, also generated a backlash: after crime reached its peak in the 1980s, there were widespread demands for more police presence in inner-city communities and tougher penalties for offenders.

We changed that by putting many more police officers on the streets, and crime did fall. But that doesn’t mean stop-and-frisk was the source of the drop. Experimental and quasi-experimental data that we have suggests that police presence (independent of stop-and-frisk) is effective. Look at any of several recent analyses of policing, and you see that implementing more foot patrols appears to reduce crime. To be sure, the NYPD put more boots on the ground, but under former Commissioner Kelly, the increased presence was accompanied by unduly aggressive practices that have come to define stop-and-frisk, leading some to believe that those invasive practices should get the credit for reducing crime.

There’s an old saying: “You can’t beat something with nothing.” So what does good policing look like? To find the answer, just look to the experts who have conducted large-scale empirical studies. Thanks to “Broken Windows” co-author George L. Kelling and his collaborators, we’ve long had evidence that moving urban officers out of their cars is a good idea. That finding was most recently confirmed by a bevy of Temple criminologists in a large scale randomized control study of foot patrols in Philadelphia.

Thanks to Tom Tyler and his collaborators, we have ample evidence that polite and respectful interactions between police and suspects encourages people to obey the law more than impolite and degrading interactions. Thanks to Lawrence Sherman (and, more recently, Anthony Braga and Brenda Bond), we have evidence that moving police from low-crime areas to higher-crime areas is a cost-effective way to reduce crime. And thanks to Andrew Papachristos, Danielle Wallace, Tracey Meares, and Jeffrey Fagan we have evidence that firm and respectful offender notifications can significantly reduce criminal recidivism in general and shootings in particular.

Crime-prevention, though, isn’t just about policing. The Crime Lab at the University of Chicago has shown the efficacy of intensive tutoring and counseling aimed at helping at-risk youth. A large-scale RCT study of lead abatement would be more costly, but would also answer lingering questions about the lead-crime link and help policymakers determine if some of the proposals advocates are pushing are worth implementing.

Police departments across the country (including the NYPD under Commissioner William J. Bratton and Mayor Bill DeBlasio) have begun implementing programs based on hard evidence derived from large-scale studies like those described above. It is disappointing that just as these officials are embracing practices supported by empirical research (not to mention by the citizens in the communities being policed), proponents of now-discredited tactics want to turn back the clock. Happily, they are rapidly becoming relics of history, not guides for the future. Evidence-based policing and crime prevention is here to stay, and we’ll all be far safer for it.