There's no evidence stop-and-frisk works, but there's burgeoning evidence that it doesn't.
The question of whether stop-and-frisk is constitutional has (for now) been settled. Whether the policy works to reduce violent crime is another question entirely. Mayor Michael Bloomberg's strongest argument for the efficacy of stop-and-frisk is that crime has fallen concurrent with the NYPD performing millions of stops in minority neighborhoods. That argument might carry water if stop-and-frisk was the only thing the NYPD was doing to combat crime, but it's not.
That's why researchers are trying to control for as many factors as possible to judge stop-and-frisk as policy. It's no easy feat, but New York magazine's Benjamin Wallace-Wells writes that two researchers are in the process of doing so, and that their preliminary findings will come as an unpleasant surprise for Mayor Bloomberg.
Richard Rosenfeld of the University of Missouri at St. Louis (whose previous research was used to debunk the "broken windows" theory of policing), is one of the two researchers. He measured the effectiveness of stop-and-frisk by using a "5,000-person census tracts, rather than 100,000-person precinct." Wallace-Wells reports that Rosenfeld's preliminary data suggests stop-and-frisk does something, but not much:
If you look at what happens to crime in a particular census tract after stops escalate, there is an effect — crime rates do get lower. But the scale of the effect, Rosenfeld says, so far looks tiny to him. "Thousands of stops are required to achieve even small crime reductions, say, a dozen or so [crimes] citywide," Rosenfeld told me.
The second unpublished study Wallace-Wells reports on is being conducted by Jeffrey Fagan of Columbia University. His findings could be far more troubling:
His study is not yet final, but so far, he said, it seems that stop-and-frisk, if properly isolated from the other components of policing, may actually increase crime rather than decrease it. If these findings hold, they might seem slightly paradoxical: It would mean that sending cops to a high-crime area reduces crime but actually having them stop people there does not. Many studies have found that the threat of arrest has a chilling effect on crime, but we don't understand very well what it is that establishes the threat of arrest in the mind of a would-be criminal: The mere visibility of cops on the street, or their specific actions. If Fagan is right, then it may soon be possible to separate these questions statistically. "It is one thing to talk about where you send police and even how many you send," Fagan told me, "it is quite another to ask what they should do when they get there."
Wallace-Wells includes plenty of caveats in his piece (the studies aren't finished, which means they haven't been peer-reviewed or accepted for publication; Rosenfeld receives money from George Soros' Open Society Institute and Fagan "consulted for the plaintiffs on the stop-and-frisk lawsuit"). Incidentally, the response from Bloomberg's office mentioned not one of the factors:
"Criminals don’t know how many stops are conducted on one block versus the next, or based on where lines are arbitrarily drawn on a researcher’s map. They know the police conduct stops in the areas where they operate. We have the lowest rate of teens carrying guns of any major city in the country — there is a reason for that."
Which takes us back to the whole reason researchers are looking into the effectiveness of stop-and-frisk. Maybe there is only "a" reason why New York has the lowest rate of teenage gun possession of any major city, and maybe that reason is stop-and-frisk. But if that's the case, where's the data? Where's the research demonstrating as much? As Wallace-Wells writes, "Scholars are looking, with increasing sophistication, in the places where you might expect a crime-reducing effect from stop-and-frisk to show up, if one existed. So far, they haven't found anything."
Top image: REUTERS/Eric Thayer