Part 1 Part 2 Part 3 Codap R-Lab

Exploring Racial Disparities in New York City's Stop-and-Frisk Policies

Part 3: Finding Patterns by Mapping Police Precincts

In the 2013 U.S. District Court case, the NYPD defended its stop and frisk practices, saying that since crimes mostly occur in Black and Latino neighborhoods, it is understandable that more Black people would be found suspicious of crimes.⁹

Let’s look at the data to evaluate the NYPD’s claim. If this claim were the only reason for the disparity in stops, we might expect people of color to have a higher likelihood of being stopped in Black and Latino neighborhoods. In Figure 3 below, we can compare the police stops to the racial distribution in each precinct.

Figure 3A: The distribution of Black people in New York City by precinct. There is a higher population density of Black people (denoted by redder colors) who tend to live in the southeast portion of the city—in Brooklyn, South Queens, and the Bronx.

Figure 3B: The number of arrests per 1000 Black people in New York City by precinct. Arrests of Black people are happening most frequently in Manhattan, Northern Queens, and Staten Island. The number of arrests per 1000 people is low in the southeast regions (where Black population density is higher) and higher in the northwest section of the city.

These two graphs show that Black and Latino people have a much lower probability of being arrested in Black and Latino neighborhoods. In the App below, you can also click on any of the precincts and see the number of arrests within each precinct. After clicking on several locations, we can see that people of color tend to have more arrests than white people in most of the precincts. A graph alone cannot prove the inequality of arrests is caused by discrimination. However, a careful examination of these graphs can help us discredit the argument that the racial distribution of the precincts caused the racial disparity in arrests.

3A: Explore the data:

Use the app to look at multiple graphics evaluating NYPD stop and frisks based upon year, race, gender, and type of crime.

To make a graph that looks like Figure 3A, select:

Year: 2005

Color Precincts by: Racial distribution for each precint: Black

Remove Precints: 22

Scale: Logarithmic

Instructors Note: Go to faculty resources to access student data

3B: Get Curious

1. What would we expect the above graphs to look like if there was a strong relationship between the racial distribution of a precinct and the number of Black people that were arrested?

2. Add Precinct 22 back into the dataset, then click on that precinct in the MAP. Notice that Precinct 22 tends to have very different characteristics than most other precincts. As such, it is often removed from the data as an outlier. Zoom into the graph and provide a brief explanation as to why Precinct 22 may be different than other locations.

3. Precinct 114 is close to the center of the city and tends to have a large number of arrests. Zoom into this location and identify a feature of this precinct that could explain a large number of arrests in this area. (Hint: What is Rikers Island?)

4. What are some practical ways that citizens, lawmakers, police departments, or other stakeholders could use this data?

Go back to Part 2

3A: Explore the data:

3B: Get Curious

Dataspace

Data Stories

Stats Games

Other Links

Questions?