As all the world now knows, Judge Shira Scheindlin has ruled that the New York City Police Department’s stop-and-frisk policy amounts to “a policy of indirect racial profiling” that violates the U.S. Constitution. But how did the she reach this conclusion? The answer turns out to be pretty interesting. It involves a number of statistical studies presented to the court by expert witnesses for the plaintiffs (a number of New Yorkers who claimed to have been stopped and frisked without cause) and the defense (the city of New York).
From my reading of Judge Scheindlin’s opinion, which runs to almost two hundred pages, and a brief perusal of some of the expert testimony, the court’s decision seems like the correct one. (Like many New Yorkers, I think the stop-and-frisk policy went way too far and, in some neighborhoods, degenerated into racial harassment.) Still, Judge Scheindlin’s ruling isn’t without its quirks. In one instance, she criticized the conclusions of the plaintiffs’ expert witness; in another, she endorsed the same expert’s work and relied upon it to justify her ruling that stop-and-frisk is indirect racial profiling. If and when the city goes forward with its appeal, the methodology of the studies, and the interpretations that Judge Scheindlin placed upon their conclusions, are sure to be central issues. (For more on Scheindlin, see the piece that Jeffrey Toobin wrote for the magazine about her and the case earlier this year.)
As I say, I think she got it right. But let’s see how she got there. Because of the vast number of stops that the N.Y.P.D. has carried out—about 4.4 million between January of 2002 and June of 2012—the court couldn’t hope to ascertain whether the police’s actions were justified in each and every case. So, in addition to examining nineteen individual cases, with testimony from those involved, Judge Scheindlin relied heavily on some statistical analysis carried out by the plaintiffs’ chosen expert, Dr. Jeffrey Fagan, a criminologist at Columbia Law School who has long taken an interest in policing issues. And the city, in turn, commissioned a critique of Fagan’s work by two other scholars: Dennis Smith, of N.Y.U.’s Wagner Graduate School of Public Service, and Robert Purtell, of the SUNY system’s University at Albany.
Fagan carried out two studies—one related to the claim that stop-and-frisk violates the Fourth Amendment’s protection against unreasonable searches and seizures, the other related to the claim that the policy, by discriminating based on race, violates the Fourteenth Amendment’s guarantee of equal protection under the law for all American citizens.
In the first study, Fagan examined the information contained on UF-250 forms that N.Y.P.D. officers are required to fill out after stopping and frisking somebody. The forms contain boxes in which the officers have to indicate why they initiated the encounter—the options included “Actions indicative of ‘casing’ victims or location,” “Fits description,” and “Suspicious Bulge/Object”—and whether it led to further action, such as an arrest. (Fagan didn’t actually see the completed forms, but he had access to an electronic database that recorded their contents.) After examining the data from the UF-250s, Fagan categorized each encounter as “apparently justified,” “apparently unjustified,” or “ungeneralizable.” On this basis alone, he concluded that six per cent of the stops—about two hundred thousand of them—were “apparently unjustified.”
In another statistical exercise, Fagan compared the number of stops in each enforcement area, and the race of the people stopped, to a benchmark that he constructed based upon the racial composition of the area and its crime rate. This benchmark was designed to provide a rough guide to what the racial distribution of stops would have been if the police officers carrying them out had been acting in a racially neutral manner, without any personal or institutional biases. By comparing the actual history of stops in each area to the benchmark, Fagan was able to show that the number of blacks and Hispanics stopped appeared to be excessive, even allowing for the fact that these groups are more likely than others to live in high-crime areas. To be more precise, after carrying out a regression analysis, Fagan concluded:
NYPD stops are significantly more frequent for Black and Hispanic citizens than for white citizens, after adjusting stop rates for the precinct crime rates, the racial composition and other social and economic factors predictive of police activity. These disparities are consistent across a set of alternate tests and assumptions.
Blacks and Latinos are more likely to be stopped than Whites even in areas where there are low crime rates and where residential populations are racially heterogeneous or predominantly White.
And that was not all. Fagan’s analysis also showed that blacks and Hispanics, once they had been stopped, were more likely to be subjected to the use of force, even though the probability of the stop resulting in further action—like an arrest, or a summons—was actually lower in cases involving minorities than in those involving whites.
The city and its expert witnesses, Smith and Purtell, seized upon some of Fagan’s findings and tried to discredit others. With regard to the Fourth Amendment claim, they argued that because Fagan concluded, based upon the UF-250s, that just six per cent of the stops were “apparently unjustified,” the plaintiffs had failed to show that the N.Y.P.D. had a policy of stopping people without reasonable suspicion that they might be doing something illegal, or preparing to.
On the second issue, of equal protection and racial profiling, Smith and Purtell argued that Fagan’s benchmark for race-neutral behavior was misleading, and they constructed one of their own. Rather than using data indicating the racial composition of local areas where the stops were carried out, and their levels of crime, the academics relied on data about the race of crime suspects, as derived from the descriptions of crime victims, and the race of people arrested. Defending the fact that close to nine in ten of the people stopped under the program in 2011 and 2012 were black or Hispanic, the city pointed out that “approximately 83% of all known crime suspects and approximately 90% of all violent crime suspects were Black and Hispanic.” Based on their analysis, Smith and Purtell argued that the city wasn’t racial profiling. Rather, they argued, the disproportionate stopping and frisking of blacks and Hispanics could be attributed to the disproportionate number of black and Hispanic people in the pool of criminals and potential criminals.
In assessing these claims and counterclaims, Judge Scheindlin made clear that she regarded “Dr. Fagan as a more reliable expert than Drs. Smith and Purtell.” All the same, she took issue with some of his findings, pointing out that he was “extremely conservative in characterizing stops as lacking reasonable suspicion.” For one thing, she pointed out, the information on the UF-250 forms was often unreliable (officers frequently fail to fill out the forms properly, or at all) and one-sided (the forms record only the police’s account of what happened). Moreover, Scheindlin pointed out, many of the justifications that the officers ticked for making a stop, such as “Furtive movements” and “High crime area,” were so vague and subjective as to be practically meaningless. “In light of Dr. Fagan’s very generous assumptions in categorizing the stops, his analysis can best be understood as providing a very rough minimum number of unjustified stops,” Judge Scheindlin wrote. “The actual number of unjustified stops was likely far higher.”
When it came to the competing statistical studies of alleged racial profiling, each of them based on a different standard for race neutrality, the judge was equally clear. “I conclude that Dr. Fagan’s benchmark is the better choice,” she wrote. The reason, she explained, was that “Fagan’s benchmark captures what the NYPD’s stops would look like in the absence of racial discrimination: his use of local population data reflects who is available to be stopped in an area (assuming, as the evidence shows, that the overwhelming majority of stops are not of criminals), and his use of local crime rates reflects the fact that stops are more likely to take place in areas with higher crime rates.”
In contrast, the benchmark that Smith and Purtell constructed, based upon the races of criminal suspects, was misleading, Scheindlin said. She maintained that this was the case because most of the people stopped weren’t really suspects at all: they were law-abiding citizens—as evidenced by the fact that, in nearly nine out of ten cases, they were allowed to proceed without being arrested or issued a summons. “There is no basis for assuming that the racial distribution of stopped pedestrians will resemble the racial distribution of the local criminal population if the people stopped are not criminals,” Scheindlin wrote. And given the fact that most of them aren’t criminals, or people who fit the description of criminals, stop-and-frisk amounted to unconstitutional “racial profiling based on local criminal data.”
As I said, it’s all very interesting, as well as very important. To be sure, some of the issues involved have been raised before. In 1999, Fagan was one of the academics who worked on a study of stop-and-frisk for the state attorney general, who was then Eliot Spitzer, which reached conclusions that were similar to the ones Judge Scheindlin drew. (Sadly, nothing much came of that report.) Since then, the statistical techniques have evolved. In constructing his test for race neutrality, Fagan used a method similar to one that Ian Ayres, a professor at Yale Law School, employed in a 2008 study of pedestrian and motor-vehicle stops carried out by the Los Angeles Police Department.
Now, though, the stakes are higher than ever. In explicitly basing at least part of her ruling on complicated statistical analyses that most ordinary jurors wouldn’t understand, Judge Scheindlin has raised anew the question of what sort of weight should be put on academic testimony in cases of this sort. If Mayor Bloomberg’s promised appeal goes ahead, the city will doubtless hire some new number-crunchers to have another crack at defending stop-and-frisk, and at undermining Fagan’s analysis. The “battle of the experts”—that’s another phrase Judge Scheindlin used—will continue, possibly all the way to the Supreme Court.
Photograph by Allison Joyce/Getty.
Judge of Federal District Court in New York upheld the bedrock principle of individual liberty on Monday when she ruled that the tactics underlying ’s program violated the constitutional rights of minority citizens. She found that the city had been “deliberately indifferent” to police officers illegally detaining and frisking minority residents on the streets over many years.
Judge Scheindlin was clearly speaking of Mayor when she concluded: “The City’s highest officials have turned a blind eye to the evidence that officers are conducting stops in a racially discriminatory manner. In their zeal to defend a policy that they believe to be effective, they have willfully ignored overwhelming proof that the policy of singling out “the right people” is racially discriminatory and therefore violates the United States Constitution.”
The judge made clear that she was not striking down the program — which remains an important tool for law enforcement — but requiring the city to use that tool in a way that does not discriminate against African-Americans and Hispanics and that comports with constitutional guarantees against unreasonable search and . Given the city’s refusal to alter its practices significantly, Judge Scheindlin had little choice but to appoint an outside monitor to oversee sweeping changes in how the trains its officers and carries out the stop-and-frisk policy.
Under the Fourth Amendment, police officers can legally stop and detain a person only when they have a reasonable suspicion that the person is committing, has committed or is about to commit a crime. Over the years, however, the Police Department has adopted a strategy that encourages cops to stop and question mainly minority citizens first and to come up with reasons for having done so later. This has resulted in people in some neighborhoods being stopped without reason scores of times a year. These unconstitutional stops, Judge Scheindlin wrote, have exacted a “human toll” in demeaning and humiliating law-abiding citizens. She is currently overseeing three lawsuits against this troubled program. The ruling issued on Monday, in Floyd v. The City of New York, was filed by plaintiffs alleging racial profiling in street stops.
At the heart of the Floyd case are statistics showing that the city conducted an astounding 4.4 million stops between January 2004 and June 2012. Of these, only 6 percent resulted in arrests and 6 percent resulted in summonses. In other words, 88 percent of the 4.4 million stops resulted in no further action — meaning a vast majority of those stopped were doing nothing wrong. More than half of all people stopped were frisked, yet only 1.5 percent of frisks found weapons. In about 83 percent of cases, the person stopped was black or Hispanic, even though the two groups accounted for just over half the population.Continue reading the main story