Summary of Traffic Stops Statistics
Raleigh Police Department
The following report compiles and presents statistical summaries of traffic stops statistics drawn from the NC State Bureau of Investigation database to which each police agency reports their statistics each month. Data in this report cover the period of 2002 through 2020. The data exclude checkpoint stops because such stops are not recorded systematically. The data also include only the driver of the vehicle, excluding any passengers. Passenger information is generally recorded only in the event of an adverse outcome (e.g., search, arrest).
This report provides the following summary statistics:
First is a table providing summary statistics on numbers of stops, searches, contraband hits, and arrests, as well as relevant rates of these outcomes. This is followed by a series of graphics displaying department and officer level data.
| Summary of outcomes | State-wide | RPD |
|---|---|---|
| Stops | 24,980,776 | 1,045,289 |
| Traffic Safety Stops | 13,365,910 | 451,793 |
| Searches | 763,343 | 37,836 |
| Hits | 280,152 | 7,545 |
| Arrests | 500,040 | 15,467 |
| Arrest From Hit From Search | 105,784 | 2,628 |
| Consent Searches | 346,475 | 9,808 |
| Arrest From Hit From Consent Search | 20,759 | 248 |
| Probable Cause Searches | 264,963 | 15,176 |
| Arrest From Hit From Probable Cause Search | 54,326 | 1,770 |
| Percent Traffic Safety Stops | 53.50% | 43.22% |
| Search Rate Per Stop | 3.06% | 3.62% |
| Hit Rate Per Search | 33.48% | 19.13% |
| Arrest Rate Per Hit | 38.85% | 35.03% |
| Hit-and-Arrest Rate Per Search | 13.86% | 6.95% |
| Hit-and-Arrest Rate Per Probable Cause Search | 20.50% | 11.66% |
| Hit-and-Arrest Rate Per Consent Search | 5.99% | 2.53% |
| Arrest Rate Per Stop | 2.00% | 1.48% |
| Hit-and-Arrest Rate Per Stop | 0.42% | 0.25% |
Numbers at top of the table show raw values; numbers below are the percentages based on the numbers above.
Figure 1 shows the total number of stops per year from 2002-2020. Numbers range from approximately 20,000 (in 2020) to 90,000 (in 2010). There is a drop associated with the global pandemic in 2020.
Figure 2 shows the total number of traffic stops by race and gender from 2002-2020. For example, in 2002, roughly 30% of total traffic stops were of white males, roughly 26% of stops were of black males, roughly 18% were of white females, and roughly 15% were of black females. White male driver’s share of those stopped decreased over the timeframe, while both black males and females slightly increased as a proportion of total stops. From 2002-2007, white males constituted a larger share of stops than black males, however since 2008, black males constitute the largest percentage of stops. Black females also surpassed white males in 2017 until 2020, and this is largely due to the decreasing percentage of white male stops rather than any significant changes in the rates of black females.
Figure 3. Percent of stops for “safety” purposes (speeding, stop sign, DUI, unsafe movement)
Figure 3 shows the percentage of traffic stops that are considered safety-related for each respective race and gender grouping, from 2002-2020. Safety-related spots are composed of stops due to speeding, stop sign violations, DWI, and unsafe movement. Stops for violations other than this are classified as non-safety related. These are composed of non-moving violations such as equipment violations and expired tags. For this reason, non-safety-related stops can be considered to be used as informal criminal investigations. Each race and gender grouping follows a similar trend from 2002-2020. White females and white males have a similar percentage of safety-related stops throughout the time period. Black drivers consistently had a lower percentage of their stops classified as safety-related as compared to white drivers.
Figure 4. Percent of stops resulting in a search, by race-gender category
Figure 4 shows the percent of traffic stops that result in a search for each respective race and gender grouping, from 2002-2020. The Figure indicates that most traffic stops do not result in a search. For all categories, drivers were searched at less than a 10% rate across the time period. Black males were typically searched at higher rates, ranging from 5 to 10% across the time period. This is roughly two to three times as often as the second most searched group, white males, who were searched in 2 to 4% of stops. White and black female drivers experienced similar levels of search rates over the time period, with both groups generally being searched at a lower rate than white and black males, with the lone exception of 2016 to 2019, when black females were searched at a higher rate than white males.
Figure 5. Percent of stops resulting in arrest, by race-gender category
Figure 5 shows the percent of traffic stops that result in an arrest for each respective race and gender grouping, from 2002-2020. Across the board, arrest rates are consistently very low, at under 5% for all categories across the time period. Black males are arrested at the highest rate, and white males are arrested at the second highest rate. Females, regardless of race, are arrested at lower rates than males. All race and gender groups experienced a sharp increase in arrest rates in 2014, and the rates after this year are generally higher than the rates preceding it. From 2002- 2006, arrest data is missing and therefore is not displayed in the Figure.
Figure 6. Percent of searches resulting in contraband, by race-gender category
Figure 6 shows the percent of searches that result in the finding of contraband for each respective race and gender grouping, from 2002-2020. Across the time period, there is a clear upward trend in contraband hits per search. In 2002, the rate of contraband discovery was between 8% and 18% for all categories, and this steadily increased until 2018, in which contraband was discovered between 20% and 40% of searches (2019 and 2020 showed further increases for the race-gender groupings which had sufficient data in these years, as well). The percent of searches leading to contraband is extremely similar for each of the race and gender categories, increasing at a similar rate across the time period. Note that there is insufficient hit rate per search data for various categories at different points across the time period.
Figure 7. Percent of searches resulting in contraband and arrest, by race-gender category
Figure 7 shows the percent of searches that resulted in the discovery of contraband and an arrest for each respective race and gender grouping, from 2002-2020. Note that in many instances, contraband hits do not lead to arrest due to it being found in small amounts. As seen in the Figure, the rate of contraband hit and arrest per search has experienced significant variation across the time period, with generally no clear trend. Contraband and arrest rates from searches of all drivers fell between 4% and 15%, with the sole exceptions being black females falling below 4% in 2012 and 2013, and white males exceeding 15% in 2019. From 2002-2006, arrest data is missing and therefore is not displayed in the Figure. Note that there is also insufficient data for various categories at different points across the time period.
Figure 8. Percent of stops resulting in contraband and arrest, by race-gender category
Figure 8 shows the percent of stops that resulted in the discovery of contraband and an arrest for each respective race and gender grouping, from 2002-2020. The rates are very low for each category. White males, white females, and black females all have rates of less than 0.5% across the time period with very little variation. The rates for black males have increased slightly. From 2007 to 2013, the rate hovered around 0.5%, but after a sharp increase in 2014, the rate of contraband and arrest in stops of black males has increased to between 1% and 1.5% year to year. Overall, for all race and gender categories, traffic stops are overwhelmingly unlikely to result in contraband and arrest. From 2002-2006, arrest data is missing and therefore is not displayed in the Figure.
Figure 9. Black-White Ratio of search rates and arrest rates, for males
Figure 9 compares the search rate and arrest rate of black males and white males from 2002-2020. The search rate ratio is the search rate of black males divided by the search rate of white males, while the arrest rate ratio is the arrest rate of black males divided by the arrest rate of white males. A search rate ratio of 1.0 would indicate that black and white males are both searched in an equal percentage of their respective traffic stops. This similarly applies for the arrest rate ratios. Across the time period, the search rate ratio is consistently at or above roughly 1.5, peaking at greater than 3.0 in 2016 and 2020. This means that black males are searched much more frequently than white males. The arrest rate ratio is lower than the search rate ratio for every year recorded across the time period. This means that black males are arrested at a greater rate than white males, but not to the degree that they are disproportionately searched. From 2002-2006, arrest data is missing and therefore there is no arrest rate ratio for these years.
Figure 10. Black-White Ratio of search rates and arrest rates, for females
Figure 10 compares the search rate and arrest rate of black females and white females from 2002-2020. The search rate ratio is the search rate of black females divided by the search rate of white females, while the arrest rate ratio is the arrest rate of black females divided by the arrest rate of white females. A search rate ratio of 1.0 would indicate that black and white females are both searched in an equal percentage of their respective traffic stops. Across the time period, the search rate ratio ranges from 0.7 to 1.9. The arrest rate ratio is lower than the search rate ratio for each year across the time period, although in most years the deviation is not extreme. From 2002-2006, arrest data is missing and therefore there is no arrest rate ratio for these years.
Figure 11. Summary of stops by hour over the week: Stops, percent safety-related, percent ending in citation, search, and arrest
Figure 11 displays the number of traffic stops, the percent of safety related stops, the percentage of drivers receiving a citation, the percentage of drivers searched, and the percentage of drivers arrested across the hours of the week. The Figure indicates that there is not much variation between the days of the week in regard to these measures, and instead shows that the variation is seen within the hours of the day. There are a few noticeable trends throughout the week, however. Arrest rates are slightly higher in the very early morning (directly after midnight) on Fridays, Saturdays and Sundays.
Figure 12. Summary of stops by hour over the day: Stops, percent safety-related, percent ending in citation, search, and arrest
Figure 12 displays the number of traffic stops, the percent of safety related stops, and the percent of drivers searched, given a citation, and arrested by hour of day. There are a number of clear trends indicated in the Figure. The frequency of a traffic stop is lowest in the early hours of the morning, from around 4 to 6AM. The number of traffic stops increases until 8AM, at which the frequency is similar until around 3PM. From 3PM to 6PM stops decrease, and stops then increase until 11PM. At midnight, the frequency of stops begins to drop.
The percentages of drivers searched and arrested both have a similar trend across the time of day. Drivers are both searched and arrested at the highest rate in the few hours after midnight, and then these rates of search and arrest decrease dramatically until 5AM. After 5AM, the rate of both search and arrest is at extremely low percentages, and both slowly increase throughout the daytime and into the late night time hours.
The percentage of drivers receiving a citation follows an opposite trend of search and arrest rates. The rate of citations given is at its highest at 8AM, and it stays at this point for much of the daytime hours. At roughly 5PM, citation rates begin to decrease steadily until the nighttime hours and the very early morning, at which citation rates are the lowest. The rate is significantly higher during daytime hours as compared to nighttime hours.
The percent of safety-related stops varies throughout the day, reaching its highest point at 5AM and then steadily decreasing for the rest of the day until a minimum is reached at 7PM. From 7PM to 5AM the rate increases, with dramatic increases in the percentage of safety-related stops indicated from 3AM to 5AM.
Figure 13. Hourly breakdown of percent of stops by race-gender category
Figure 13 displays the share of overall stops represented by the respective race and gender categories by hour of day. The Figure indicates that black males are more likely to be stopped during the nighttime hours directly before and after midnight than they are to be stopped during the daytime. At 4AM, black males reach a high of 40% of those that are pulled over, and by 7AM their share of stops decreases to 20%, and stays at a lower proportion of stops throughout the daytime hours. White males, in contrast, make up a relatively consistent share of the overall traffic stops throughout the day, at around 20 to 25%. There are no significant variations during the daytime or nighttime. Black females make up a relatively consistent share of traffic stops throughout the day, with similar levels to white males. White females show an increase in their likelihood of being stopped during the daytime hours, during which they are a higher percentage of stops than black females. During the hours 6PM to 7AM, the opposite is true.
Figure 14. Hourly breakdown of percent of stops related to traffic safety, by race-gender category
Figure 14 displays the percentage of stops that are for safety-related purposes for each race and gender category, by hour of the day. Safety-related purposes refer to traffic stops that are meant to prevent moving violations, and effectively keep the road safe. Stops that are not safety-related are often used for investigatory purposes, in which case the officer is generally seeking to conduct an informal criminal investigation of the driver. As evidenced in the Figure, black drivers are much less likely to be stopped for safety-related purposes than white drivers throughout the course of the day. In turn, this means that black drivers are more likely to be subjected to investigative stops that are not for moving violations. These stops are most common during the nighttime hours, when the percentage of safety-related stops is lowest. At any given hour of the day, black drivers are more likely to be stopped for a non safety-related purpose than white drivers.
Figure 15. Hourly breakdown of percent of stops resulting in search, by race-gender category
Figure 15 displays the percentage of traffic stops which result in a search for each race and gender category, by hour of the day. Black males are searched at the highest rate throughout the day, with the exception of 2AM when white males are searched at a marginally higher rate. The time of day does significantly influence the rate at which they are searched. Approaching and during the nighttime hours before and after midnight, from approximately midnight to 4AM, black males are most likely to be searched during a traffic stop, and this search rate decreases during the daytime hours. This pattern applies for all of the gender and race categories. During daytime hours, however, black males are still significantly more likely to be searched than other race and gender categories. White males, white females, and black females are searched at similar rates throughout the daytime hours and into the nighttime hours before midnight. The Figure also indicates that a majority of stops do not result in a search regardless of race and gender category or time of day.
Figure 16. Hourly breakdown of percent of stops resulting in arrest, by race-gender category
Figure 16 displays the percentage of traffic stops which result in an arrest for each race and gender category, by hour of the day. For all categories, drivers are more likely to be arrested during nighttime hours, specifically the hours directly after midnight. Stops are most likely to result in an arrest at 2AM. During these hours directly following midnight, white drivers are more likely to be arrested than black drivers. However, black males are more likely to be arrested than all other drivers from 4AM until 11PM, which constitutes a large majority of the day. The Figure also indicates that a large majority of stops do not result in an arrest regardless of race and gender category or time of day.
Figure 17. Number of stops by officer
Figure 17 displays the distribution of number of stops by officer across the 2002-2020 time period. All officers displayed have over 100 traffic stops. 1,073 of the 1,378 officers included in the data have less than 1000 traffic stops.
Figure 18. Percent black of drivers pulled over by officer
Figure 18 displays the percentage of stopped drivers that are black for individual officers, from 2002-2020. The distribution includes officers with greater than 100 traffic stops. For the median officer, 45.9% of stops are of black drivers, while black people only make up 29% of the population in the city. This means that a majority of officers are stopping black drivers at a rate disproportionate to their share of the population. This disparity between stop behavior and population data is often greater, with black drivers composing more than 75% of stops for a significant portion of officers. At the far right, the Figure shows that there are 140 officers whose mix of drivers stopped is more than 75% black.
Figure 19. Officer search rates by race
Figure 19 displays the search rates of individual officers for the given race and gender category. The criteria for an officer’s inclusion is 100 or more traffic stops, as well as at least 50 stops of the specified race and gender category, from 2002-2020. In the first graph, the median officer has a search rate of 4.5%, meaning that black male drivers are searched in 4.5% of traffic stops. Many officers have search rates that are higher than this median value, meaning that searches occur more often for certain officers in traffic stops involving black males. For white male drivers, the median officer has a search rate of 1.8%. This indicates that the median officer in Raleigh is searching black males at a higher rate than they are searching white males.
Figure 20. Black-White Ratio of search rates by officer, for males
Figure 20 displays the distribution of black-white male search rate ratios across the officers which meet the criteria. The criteria for an officer’s inclusion is 50 traffic stops of both white males and black males, from 2002-2020. The “Black-white male search rate ratio” can be interpreted as an officer’s search rate of black male drivers divided by their search rate of white male drivers. An racially equitable outcome would therefore be 1, meaning that black and white male drivers are searched in an equal percentage of traffic stops. The median officer instead has a search rate ratio of 1.77, meaning that the median officer searches black male drivers at a higher rate than white male drivers. A significant number of officers have search-rate ratios that are much higher than the median, with some searching black male drivers at 3 or more times the rate of white male drivers.
Figure 21. Search rate versus hit rate by officer
Figure 21 presents the hit rates and search rates of individual officers who meet the criteria. As evidenced by the low correlation, there is great variability in both the propensity of individual officers to search drivers and their success in finding contraband. One might expect that when officers have a high search rate but a low hit rate, that their supervisors would advise them to adjust their search rate in order to avoid so many fruitless searches. Similarly, for officers with low search rates but high hit rates, one might expect supervisors to instruct them to conduct more searches because they are being too cautious. Such a learning process would generate relative consistent hit rates; where officers are too high, they would be advised to do more searches, and where too low, to conduct fewer. The Figure shows clearly that this is not happening. Search and hit rates range from low to high, with a very low correlation between the two. Certain officers search drivers at extremely high rates but have very low hit rates, meaning that their threshold of suspicion is likely too low. Other officers search at a much lower rate and have very high hit rates, showing that their threshold of suspicion may be too high. Overall, the Figure shows that officers are not gravitating towards a single common range of hit rates. This suggests that the department does not seem to hold its officers to a common standard of search rate success. Rather, each officer decides for him or herself how aggressive to be in searching, with virtually no guarantee that more searches will lead to lower hit rates, or that fewer searches will be targeted on those most likely to have contraband.