Better for everyone: Black descriptive representation and police traffic stops

ABSTRACT

Racial disparities in citizen interactions with police are ubiquitous concerns in American communities. What difference does electoral representation make? We demonstrate that black descriptive representation in local government affects police activity and scrutiny in a given community. We use a new dataset comprised of over 79 municipal police departments spanning 6 states, based on tens of millions of individual-level traffic stops. In cities and towns with majority-black city councils, traffic stops are less likely to result in a search. This decline in search rates affects both white and black drivers, though the decline is larger for black drivers. Even after controlling for socioeconomic factors, segregation, and crime rates, descriptive representation still matters. A city council composed of a majority of black members is associated with important differences in policing, affecting both white and black residents.

KEYWORDS:

• policing
• racial disparities
• traffic stops
• descriptive representation
• local politics

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 The purpose of imposing a threshold is to ensure that the search rates we calculate, by agency, are robust (i.e., that they will not change drastically with the introduction of one or two additional searches or stops). This is especially important because searches after a traffic stop are relatively rare – meaning rates can be subject to fluctuations. We used this threshold to determine the municipalities for which we would collect political variable data.

2 All results hold when we use the black–white search rate difference, another measure of inequality. See Appendix C.2 for these analyses.

3 Note that we consider the possibility that a majority of black members may not be necessary to affect policing. Rather than using a dichotomous variable, we instead take the proportion of city council seats held by black members, and use this to predict search rates and disparities. These models are presented in Appendix C.1. The results are largely null. This indicates that a majority is necessary before changes are seen. We also consider whether the proportion of black members may affect policing, conditional on whether there is a majority. We again largely find that it is more about having a majority than about the proportion itself (though we do find this interaction to be significant for the search rate of white drivers). See Appendix C.1 for this analysis.

4 Demographic and municipal information comes from the US Census Bureau and American Community Survey. Race of the police chief is original data, collected by the authors. The proportion of the police force that is black is the mean level, calculated from the 2000, 2003, 2007, and 2011 Law Enforcement Management and Administrative Services survey. See Appendix A for more detail on these measures.