On a first glance at the data, we can see that this data covers traffic stops by police officers in Charlotte, North Carolina over the year 2016. Let’s explore first the race, gender and years of service of officers.

Officer Race Officer Gender Count
White Male 45,939
Black/African American Male 11,708
White Female 4,861
Asian / Pacific Islander Male 3,731
Hispanic/Latino Male 2,198
Black/African American Female 606
Not Specified Male 338
Hispanic/Latino Female 321
American Indian/Alaska Native Male 221
Not Specified Female 90
Native Hawaiian/Oth Pac Island Male 62

We omit the NAs in the data, because we can see that there are 261 NAs in the data. Also, there are unspecified data for officer races of 460 officers.

We observe that most of the stops were being conducted by white male officers, followed by black/African American officers, followed by white female officers. Asian and Hispanic/Latino officers conducted the least number of stops during 2006.

Officer Race Gender Average Years of Service
Asian / Pacific Islander Male 12
White Male 12
Black/African American Male 11
American Indian/Alaska Native Male 9
White Female 9
Black/African American Female 8
Hispanic/Latino Male 7
Native Hawaiian/Oth Pac Island Male 7
Hispanic/Latino Female 4


Now, let us summarise and visualize the officer data by their respective CMPD divisions.

CMPD Division Count
South Division 8,699
Providence Division 7,634
North Division 6,782
Westover Division 6,642
Hickory Grove Division 6,383
Eastway Division 5,833
University City Division 5,302
Steele Creek Division 4,788
Independence Division 4,613
Metro Division 3,652
Central Division 3,148
Freedom Division 3,130
North Tryon Division 3,041

We observe that there is a high number of NAs in the CMPD Division data of officers, which is 9586. We also observe that the highest number of officers were from South Division, Providence Division, North Division, Westover Division and Hickory Grove Division respectively in decreasing numbers.

Let us know visualize the officer data that we have previously summarized.

We observe that the highest number of stops were being conducted by officers of white race from the “South Division”, followed by Providence Division and Hickory Grove Division. A lot of stops were also being conducted by black/African American officers.

Let us now visualize the number of stops by months of 2016.

We observe that the maximum number of stops occurred for vehicle regulatory and speeding reasons.

Now, let us investigate the result of the stops as well as the race/ethnicity and age of the driver.

We observe that the majority of drivers which were stopped in 2016 were Non-Hispanic Black and Non-Hispanic White. We also observe that the result of the stop was a verbal warning followed by citation issued. This remained the same for all racial and ethnic groups of people. Let us further dive into the age distribution of these drivers, and whether a search was conducted by the officers.

We can observe from the boxplot above that the drivers belonging to the race and ethnicity group of Non-Hispanic White and Non-Hispanic Black have the highest amount of outliers with respect to age. We investigate whether a search was conducted for these drivers.

Driver Race and Ethnicity Search Conducted Number
Non-Hispanic Black No 40,365
Non-Hispanic White No 25,882
Non-Hispanic Black Yes 2,467
Non-Hispanic White Yes 464

We observe that a search was not conducted for 94.24% of Non-Hispanic Black drivers, and for 98.24% of Non-Hispanic White drivers, which is a significant percentage of the drivers. We also notice a search was conducted for 5.76% of Non-Hispanic Black drivers, which is significantly higher than 98.24% of Non-Hispanic White drivers.

Let us now explore the reasons of stops which occurred for drivers of Non-Hispanic White and Non-Hispanic Black race and ethnicity. Let us also explore the officers which conducted these investigations.

Thus, we observe that the major reasons for stopping for Non-Hispanic White and Non-Hispanic Black drivers were also vehicle regulatory and speeding. Thus, there is no major outlier in this data.

Repository