Association for Public Policy Analysis & Management

Incarceration rates for black males are almost three times higher than for whites (Blumstein 2001). Existing research suggests that a main driver of racial disparities in arrest rates is racial profiling and random searches of black men in high poverty neighborhoods by police (Alexander 2009, Blumstein 2001, Rios 2011). Yet, contemporary policing in the US is largely reactive, meaning police respond to calls for service from community members, rather than officer-initiated. To my knowledge, no research has evaluated the role of calls for service to the police in contributing to disparities in arrest rates. I examine several research questions: What proportion of arrests stem from calls for service to the police versus officer-initiated stops? What types of calls are more likely to result in arrests? Are arrests more likely to be made when calls comes from neighborhoods with higher proportions of African Americans? Answers to these questions will be useful in understanding additional mechanisms that influence arrest rates and in rethinking how call-centers filter calls for service.

To address these questions, I analyze call for service and arrest data from Washtenaw County, Michigan. The data include 19,124 calls for service from residents over a one year period and over 1,400 arrests. I merge American Community Survey 5-year estimates from 2013 at the census tract level to the 2011 call for service and arrest data to explore relationships between calls and demographic characteristics of census tracts. I focus on Ypsilanti Township which has a population of 54,719 and is largely suburban with considerable racial and economic heterogeneity. I supplement the call for service and arrest data with ethnographic field notes from working as a 9-1-1 communications operator at a local call center in Southeastern Michigan.

Descriptive analyses, density maps, and logit regressions (with the outcome being arrest or not) suggest that policing in a democratic society where citizens dictate where and when police are activated may contribute to disparities in the amount of police presence in communities. I find that calls for service across Ypsilanti Township were made at a rate of 35 per 100 residents in 2011, with considerable variation across census tracts of varying racial and socioeconomic compositions. Arrest data indicate that over 80 percent of arrests stemmed from a call to the police. I am currently in talks with the Washington DC police department to determine if these findings are generalizable to larger cities. While recent events in Ferguson, MO and other cities across the US raise alarms that law enforcement engages in race-based discrimination when they police proactively, my findings suggest that policy reforms must also consider how residents themselves shape police presence in their communities and how police policies and practices for filtering and responding to citizen calls can and should mediate resident influence.