Chicken Little, Three Blind Men and an Elephant, and “Racial Profiling”: A Commentary on the Collection, Analysis, and Interpretation of Traffic Stop Data

Abstract

ABSTRACT This paper discusses the collection, analysis, and interpretation of police traffic stop data. The focus primarily is on the theoretical, statistical, and practical determination of whether law enforcement is engaging in “racial profiling.” One of the most controversial components of research in this area relates to the “baseline” or comparison group. Researchers struggle with determining whether agencies are engaging in “racial profiling” when making traffic stops without some measure of what the world would look like in the absence of “racial profiling.” We know what is, but how can we know what should be? A model is proposed that delineates between populations and subpopulations of drivers. In addition, arguments are presented that current proxy measures are inadequate as “baselines,” and that a careful determination of the appropriate population for comparison is essential in reaching any conclusions about the profiling practices of any department. Further, it is argued that data from a population of stops within any agency become the baseline for comparison through the use of multivariate (CHAID) analysis, and through comparisons to future data collected by that agency.