Statistics fight against discrimination in court


  • Author: Carlos Gómez Grajales
  • Date: 30 August 2013
  • Copyright: Image appears courtesy of iStockPhoto

A few days ago, a court officially ruled that the New York City Police Department’s stop-and-frisk program amounts to “a policy of indirect racial profiling” that violates the U.S. Constitution. This stop-and-frisk policy is a resource that authorizes police officers in New York City to stop and question thousands of pedestrian annually and search them for weapons, drugs or any illegal stuff. The only requisite is that the officers are obliged to fill a report stating the reason why a particular individual was stopped, though the forms allow for some really vague and subjective reasons, such as “Furtive movements” witnessed by the officer or the mere fact that the pedestrian was traversing a “High crime area”.

Although the nature of the program itself seems dangerously misguided through racial profiling, to actually rule it illegal, Judge Shira Scheindlin had to rely in some very interesting statistical analysis that were presented as the main evidence against the stop-and-frisk policy. The plaintiffs’ chosen expert, Dr. Jeffrey Fagan, criminologist at Columbia Law School, conducted two different statistical studies that sketched an accurate landscape of what this program had become.

thumbnail image: Statistics fight against discrimination in court

The first study presented by Dr. Fagan as evidence consisted on a detailed study of the UF-250 forms that N.Y.P.D. officers are required to fill out after they have stopped somebody. As previously mentioned, these forms included a section where the officer was supposed to state the reasons that led him to stop a pedestrian. Fagan noticed that the subjective reasons normally given couldn't always be qualified as a justifiable threat. After examining the data from electronic versions of the forms, Fagan categorized each encounter as “apparently justified”, “apparently unjustified”, or “ungeneralizable”, based on whether the officer completed the form with enough information to conclude that the examination was based on serious evidence. This first study concluded that six per cent of all the stops, which amount to about two hundred thousand, were “apparently unjustified.”

A second analysis devised by Fagan tried to verify whether minorities were being chased by the program more than they should. He did this in a similar manner to how the Chi-Square test compares the expected frequencies of a contingency table where rows and columns are independent to the actual one the researcher has, in order to test for relationships. Fagan compared the official number of stops in each area of the city, and the race of the people stopped, and compared this to the expected number of stops that he assumed would occur based upon the racial composition of the area, its crime rate and assuming that no racial discrimination was taken place. The results indicated that the NYPD stops are significantly more frequent for Black and Hispanic citizens than for white citizens, even after adjusting stop rates for the precinct crime rates and even after considering different plausible scenarios.

This being a lawsuit, the city presented its own experts that tried to refute the plaintiffs' evidence: Dennis Smith, of N.Y.U.’s Wagner Graduate School of Public Service, and Robert Purtell, of the SUNY system’s University at Albany. Regarding the first result, they argued that a 6% of "apparently unjustified" detentions were indeed a really low number, hence proving the efficiency of the program. But the judge noticed, in a surprising manner, that Fagan's estimation would be a kind rough minimum, considering the way the forms are filled and the fact that in many cases the officers forget to complete them.

The city experts also tried to discredit the plaintiffs' second statistical exercise. Here, the argument was that Fagan's number of "expected stops", considering a neutral police work was wrong. Smith and Purtell argued that a better benchmark would require data about the race of crime suspects, as derived from the descriptions of crime victims, and the race of people arrested. Since “approximately 83% of all known crime suspects and approximately 90% of all violent crime suspects were Black and Hispanic", according to their results, the larger number of stopped persons from these minorities was only due to the higher number of Hispanic and black people that are criminals. Luckily, the judge found the catch. She argued that using the races of criminal suspects as the benchmark for comparison was misleading. The reason is that most of the people stopped weren’t really suspects at all: they were law-abiding citizens. Just consider that in nearly nine out of ten times, they were allowed to proceed without being arrested or issued a summon.

In the end, what I would consider the more reliable and robust statistics endured. The judge ruled the stop-and-frisk policy illegal and ordered several changes in its execution, including the possible use of cameras to record the activity of the policemen that enforce this measure. Even though the use of statistics in court may be as old as the discipline, what I found refreshing is the fact that the court not only relied in the statistics but it actually looked for possible flaws in the numbers that would point to the more accurate results. And even if the plaintiffs' statistics still require further development, the court positively considered that all the evidence signaled a problem that required intervention. Once again, statistics saved the day.


Related Topics

Related Publications

Related Content

Site Footer


This website is provided by John Wiley & Sons Limited, The Atrium, Southern Gate, Chichester, West Sussex PO19 8SQ (Company No: 00641132, VAT No: 376766987)

Published features on are checked for statistical accuracy by a panel from the European Network for Business and Industrial Statistics (ENBIS)   to whom Wiley and express their gratitude. This panel are: Ron Kenett, David Steinberg, Shirley Coleman, Irena Ograjenšek, Fabrizio Ruggeri, Rainer Göb, Philippe Castagliola, Xavier Tort-Martorell, Bart De Ketelaere, Antonio Pievatolo, Martina Vandebroek, Lance Mitchell, Gilbert Saporta, Helmut Waldl and Stelios Psarakis.