[Paper Review] A large-scale analysis of racial disparities in police stops across the United States

Emma Pierson, Camelia Simoiu+5 more|arXiv (Cornell University)|Jun 18, 2017

Policing Practices and Perceptions20 citations

TL;DR

This large-scale study analyzes over 60 million traffic stops across 20 U.S. states (2011–2015) to examine racial disparities in policing. It finds that Black drivers are stopped more frequently relative to their population share, and after being stopped, are more likely to be ticketed, searched, and arrested than white drivers—evidence of a lower search threshold for Black and Hispanic drivers, suggesting bias. Legalization of recreational marijuana in Washington and Colorado reduced search and misdemeanor rates overall but did not eliminate racial gaps.

ABSTRACT

To assess racial disparities in police interactions with the public, we compiled and analyzed a dataset detailing over 60 million state patrol stops conducted in 20 U.S. states between 2011 and 2015. We find that black drivers are stopped more often than white drivers relative to their share of the driving-age population, but that Hispanic drivers are stopped less often than whites. Among stopped drivers -- and after controlling for age, gender, time, and location -- blacks and Hispanics are more likely to be ticketed, searched, and arrested than white drivers. These disparities may reflect differences in driving behavior, and are not necessarily the result of bias. In the case of search decisions, we explicitly test for discrimination by examining both the rate at which drivers are searched and the likelihood searches turn up contraband. We find evidence that the bar for searching black and Hispanic drivers is lower than for searching whites. Finally, we find that legalizing recreational marijuana in Washington and Colorado reduced the total number of searches and misdemeanors for all race groups, though a race gap still persists. We conclude by offering recommendations for improving data collection, analysis, and reporting by law enforcement agencies.

Motivation & Objective

• To assess racial disparities in traffic stops across the U.S. using a large, standardized dataset of over 60 million stops from 20 states.
• To determine whether observed disparities in stop rates and post-stop outcomes stem from bias or differences in driving behavior.
• To investigate whether search decisions reflect racial bias by analyzing both search rates and the likelihood of finding contraband.
• To evaluate the impact of recreational marijuana legalization in Washington and Colorado on stop outcomes and racial disparities.
• To recommend best practices for law enforcement data collection, transparency, and analysis to improve equity and accountability.

Proposed method

• Compiled a national dataset of 60 million state patrol stops from 20 U.S. states via public records requests, standardizing data across jurisdictions.
• Used multivariate statistical models adjusting for age, gender, time, location, and driving behavior to isolate racial disparities in stop and post-stop outcomes.
• Conducted a threshold analysis comparing search rates and hit rates (contraband found) across racial groups to detect potential bias in search decisions.
• Applied a difference-in-differences approach to assess changes in stop outcomes before and after recreational marijuana legalization in Washington and Colorado.
• Performed data quality checks, including auditing for missing values, implausible entries, and inconsistencies in racial categorizations.
• Proposed standardized data collection frameworks and recommended automated data validation and public reporting protocols for law enforcement agencies.

Experimental results

Research questions

• RQ1 Are Black drivers stopped more frequently than White drivers relative to their share of the driving-age population?
• RQ2 Do Black and Hispanic drivers face higher rates of citation, search, and arrest after being stopped compared to White drivers, after controlling for demographic and contextual factors?
• RQ3 Is the threshold for searching Black and Hispanic drivers lower than for White drivers, as indicated by lower search success rates despite similar or lower evidence?
• RQ4 To what extent did recreational marijuana legalization in Washington and Colorado reduce racial disparities in search and misdemeanor rates?
• RQ5 What data collection, reporting, and analytical practices can improve transparency and equity in law enforcement?

Key findings

• Black drivers are stopped at a rate 23% higher than White drivers relative to their share of the driving-age population, while Hispanic drivers are stopped at a rate 12% lower than Whites.
• After controlling for age, gender, time, and location, Black and Hispanic drivers are significantly more likely to be ticketed, searched, and arrested than White drivers.
• Black and Hispanic drivers are searched at higher rates than White drivers even when controlling for the presence of contraband, indicating a lower threshold for searches.
• The probability of finding contraband during a search is lower for Black and Hispanic drivers compared to White drivers, suggesting that searches of minorities are less justified by evidence.
• Legalization of recreational marijuana in Washington and Colorado reduced total search and misdemeanor rates for all racial groups, but the relative racial gap in search rates persisted.
• Data quality issues, including missing values, implausible ages, and inconsistent racial categorizations, were common, highlighting the need for improved data validation and auditing procedures.