Stanford Open Policing Project
Stanford Computational Policy Lab
Standardized traffic stop records from dozens of state patrols and cities, 31.8 million from California alone.
Police stop records gathered by public records requests and standardized into a common schema: date, location, driver race, sex, and age, reason for stop, search, contraband found, and outcome. It covers 88 locations in 42 states. The data underlie the veil-of-darkness and search-hit-rate tests for racial bias in policing.
Access
- Cost
- FreeFree download.
- License
- Open licenseOpen Data Commons Attribution License
- Sign-up
- No sign-up
- How you get it
- Bulk downloadCSV, RDS, Shapefile
What's in it
- One row is
- One row per police stop
- Covers
- 88 state patrol and city datasets in 42 U.S. states
- Years
- 2000-2020 (varies by location)
- Updated
- Irregular
- Size
- From 6,921 stops (Owensboro, KY) to 31.8 million (California statewide) per location
Quality
Careful standardization with published methods; data stop around 2020, and fields available differ by location.
- Not updated since early 2023
- Race is often recorded by the officer's perception
Connects to
- FBI NIBRS Incident Data (Crime Data Explorer): Relate stop patterns to reported crime by agency
- Mapping Police Violence: Compare everyday stops with killings by the same departments
Questions it could answer
- Does the share of Black drivers stopped fall after dark, when race is harder to see?
- Are searches of white drivers more likely to find contraband?