ABSTRACT
Identity-based stereotyping often operates on perceptions about the intersection of multiple identities. Intersectional stereotyping predicts that certain combinations of attributes lend themselves more readily to perceived suspicion than others. In this paper, I test the way that suspicion-evoking stereotypes affect police-citizen interactions. Through the use of traffic stop data from Illinois spanning ten years and amounting to more than 20 million observations, I am able to produce accurate estimates for the relative degree of targeting that individual drivers face based on their racial, gender, age, and class-based perceived identities. Overall, I find both theoretical and methodological support for the necessity of intersectional analyses of identity-based profiling.
Disclosure statement
No potential conflict of interest was reported by the author(s).
Notes
1 See Online Appendix A for more details.
2 Only eight states collect and report individual-level traffic stop data at the moment. Within those, Illinois is the only state that includes both a variable for vehicle age and for driver's age, in addition to the driver's racial and gender identity.
3 I also analyze whether or not the driver receives a ticket – the other outcome recorded in the data. The ticket analysis largely mirrors the search analysis, as it is also a harsh outcome, though there are less disparities – see Online Appendix G for a lengthier discussion.
4 Robustness checks compare this model to those estimated with different driver age cut points (21 years old and 25 years old) and different vehicle age cut points (at 3 and 5 years old) and results are nearly identical. See Online Appendix C. Further secondary analyses in the paper preserve driver age and vehicle age as continuous.