It is the first report analyzing trends in the department’s data on so-called stop-and-frisk tactics, and the fifth to be filed by the monitor, Peter L. Zimroth, to Judge Analisa Torres since a 2013 decision that found the department’s tactics to be unconstitutional.
On one hand, the report offers the most comprehensive review of recent citywide street stop activity. It shows that overall stops declined in the three years the monitor studied: to 22,563 in 2015, from 191,851 in 2013, a period in which violent crime in New York continued its decline.
Over those same years, the share of stops in which officers detained people to frisk them increased, as did the share in which officers used force, made arrests and seized guns and contraband, like drugs. As overall stops declined and fewer people were affected, officers were also stopping fewer innocent people. Making fewer, more effective, stops is one of the Police Department’s goals.
Before stopping someone, officers must suspect a crime has occurred, or is about to. The data showed a greater share of stops came after officers reasonably suspected a serious crime, such as murder or weapons possession. Less often did they stop someone for suspected low-level crimes — property crime, drugs, trespassing or so-called quality of life offenses.
The report focused on racial disparity in street stops, but not on the other question before the monitor: ensuring that the department’s stop-and-frisk practices are constitutional.
The report said assessing that question required a wider array of data as well as an analysis of the department’s progress in revising its street-stop policies and providing training on what constitutes racial profiling.
In a statement, the Center for Constitutional Rights, a plaintiff in the stop-and-frisk case, said the monitor’s report “tells us that, while some progress has been made, the N.Y.P.D. still has much work to do to end racial bias in its stop and frisk practices.”
Christopher T. Dunn, the associate legal director of the New York Civil Liberties Union, a counsel in one of the stop cases, said the drop in reported stops raised a concern that large numbers of stops were not being recorded, and that if such a failure exists, it “may be masking racial bias in stop activity.”
The Police Department did not immediately respond to the report.
Still, that a three-year band of recent data shows a broad decline in racially disparate policing is a significant turnabout from years of unconstitutional tactics during the administration of former Mayor Michael R. Bloomberg, and hews with efforts by the police commissioner, James P. O’Neill, to build trust between officers and civilians.
Recorded stops climbed steadily over most of Mr. Bloomberg’s two terms, to 601,285 in 2010, from 97,296 in 2002. They reached a height of 685,724 in 2011, but dropped and then leveled out toward the end of 2013, which was Mr. Bloomberg’s last year in office.
The monitor, Mr. Zimroth, in a letter filed to Judge Torres along with the report, expressed encouragement that racial disparities in the years analyzed “were trending in the right direction.”
The report presented several different ways of analyzing the department’s street stop data. One, cited in the federal trial leading to the 2013 decision, was a population-based analysis, while another compared recorded stops with recorded crime to derive a ratio.
But those two models led to different conclusions on whether, and by how much, officers’ initial stops of blacks and Hispanics were based on a lower degree of suspicion than stops of whites.
The first model showed that even as the estimated rate of stops in mostly black and Hispanic neighborhoods decreased between January 2013 and December 2015, residents in those places endured more stops on average than in other areas of New York.
The second model yielded different results.
By measuring stops-per-crime on each of the city’s 38,000 census blocks, it found that racial disparities ebbed substantially, citywide, over time. Of note, however, it found lingering racial disparities by comparing the average stop rates on different blocks, from different neighborhoods, with similar levels of crime.
The report outlined the weaknesses in each method, but expressed substantial confidence in the method that used stops-per-crime and bolstered the finding of diminished racial disparities in recorded stops.
Mr. Zimroth, in writing to the judge, noted the “multiple ways to slice and dice data,” and invited others to “undertake alternative analyses” in the future.
The report included one finding sure to drive further debate: It showed that even as the drops in recorded stops accelerated, the relative share of stops among racial groups remained constant. The share of stops of blacks remained about 53 percent, for Hispanics it was roughly 29 percent, and for whites it was roughly 11 percent.
“These facts do not themselves establish impermissible racial disparities because they do not speak to whether there were factors other than race that might have affected the numbers,” the report said.
The report said the factors, other than race, that affect police street stops could include crime rates in particular locations; 911 or 311 calls; and levels of street activity that could drive more encounters with the police.
The report was marked by other tidbits. For instance, it said that while both crime and street stops were geographically concentrated in the same general areas, stops were decreasing more quickly. In 2013, it said, crime in a given area was usually a good predictor of where stops would occur. Two years later, that correlation was diminished. A question will be why.
It also noted something that limited further analysis: The department has no computerized data on officer assignments that is routinely updated.
“Thus, it cannot tell us whether a higher rate of stops in minority neighborhoods was due to officers making stops of blacks and Hispanics at a lower threshold of suspicion than stops of whites and others, or due to police deployment decisions, or some combination of the two,” the report said.
Opaque deployment shrouds a variety of policies, not just stop-and-frisk, but also whether commanders assign officers based on the rise or fall of violent crime. So the monitor, for now, said he would try to use arrest and summons data to answer a fundamental question about the nation’s largest municipal police force: Where exactly are officers located around New York City?
Continue reading the main story