The present study assesses the contextual racial effects on police decisions to arrest in traffic stops.
A hierarchical logistic regression model is conducted using data collected from the Houston Police Department and US census.
The authors’ multilevel analysis indicates that the racial effect on police decision to arrest is more likely to be contextual than individual. Black and Hispanic drivers have no significant difference from White drivers in police decisions to arrest when area variations are controlled. In contrast, the concentrations of Blacks and other racial minorities in areas are significantly associated with the chance of being arrested by police in traffic stops. However, as the level of racial diversity increases in an area, the chance of being arrested is likely to decrease.
This study demonstrates the importance of racial characteristics of areas in the study of racial profiling and related police practice.
