The Ecological Sources of Racial Inequities in Proactive Policing

Abstract

This dissertation advances the thesis that racial disparities in policing are to a large degree the result of neighborhood-level processes, including organizational influences that operate at the neighborhood level. On the one hand, most research on racial disparities focuses on the influence of individual race on officer decision-making, prioritizing discretion and cognitive processes such as implicit bias and statistical discrimination. While immensely valuable, such work offers an incomplete account of racial disparities owing to the existence of various neighborhood-level influences on officer behavior. On the other hand, while there are many neighborhood-level studies of police behavior, these often invoke the same limited set of theories and use methods which do not confront major inferential challenges. As such, while research has established that place matters greatly for explaining racial disparities in policing, it remains unclear what it is about places that matter, and how much. This dissertation addresses this situation by advancing theory and evidence on the social ecological sources of racial disparities in policing in three empirical chapters. The case studied are stops made as part of the New York City Police Department’s (NYPD) as part stop-and-frisk, a proactive policing program, during its height from 2008-2012. The first empirical chapter extends research on the effect of neighborhood racial context on policing in four ways: it makes the case for conceptualizing discrimination as in part a neighborhood-level process; it synthesizes and extends theories as to how racial context may affect police behavior; it examines methodological challenges that prior work has largely failed to confront; and it takes into account all of these points to empirically examine how racial context influenced NYPD stop patterns. At the same time, the independent and interactive role of neighborhood class context is examined, and the out of place hypothesis—the idea that individual race and racial context interact to structure police behavior—is tested. Results indicate that neighborhood racial context is as powerful as individual race in structuring hit rates—a way of measuring the standards of suspicion that police typically required before stopping someone—and that this interacts with neighborhood class context to make poor black and Hispanic neighborhoods policed particularly aggressively in this sense. Neighborhood class context exerts a powerful influence on its own, independent of racial context. Conversely, however, the out of place hypothesis is not supported: police target certain racial groups in all contexts and all individuals in certain neighborhood racial contexts; the two are not found to interact in a meaningful way. Next, I turn the focus away from neighborhood racial context as measured by racial composition towards neighborhoods’ racial boundaries. That is, the emphasis is on neighborhoods’ spatial location in the larger city, particularly with respect to whether a neighborhood is surrounded by similarly raced neighborhoods versus the extent to which it borders neighborhoods of other racial contexts, that is, neighborhoods along racial boundaries. While a long line of policing scholarship has invoked the idea that policing differs at boundaries, this has been done without sustained theorizing and without direct tests of these ideas. At the same time, recent scholarship on racial boundaries for related outcomes, particularly crime and social conflict, gives added reason investigate the policing of racial boundaries, and methods with which to do so. Tests reveal that police stops are indeed more common along racial boundaries, an effect which persists on the non-black/Hispanic side (often tantamount to the white side) of divides after conditioning on a wide set of potential confounders. While social conflict as measured by 311 complaints against neighbors does not mediate this effect, much of it is accounted for by crime. However, even after accounting for crime, a racial boundary effect persists, a finding consistent with the idea that policing in part reflects the racialized use of formal social control over spaces. The final empirical chapter focuses on the role of neighborhood crime in shaping police behavior and racial disparities. Neighborhood crime is a ubiquitous control in research on racial disparities, owing to the widespread acknowledge that police focus resources and personnel where there is crime, which are also areas where minority individuals are disproportionately found, an idea known as the deployment hypothesis. However, this chapter argues that the treatment of neighborhood crime by prior literature is incomplete on two accounts. First, the deployment hypothesis is incomplete, as there are many more organizational decisions that, together, determine how powerful neighborhood crime effects will be in shaping disparities as well as the correspondence between individual and neighborhood criminality. An organizational-ecological framework is advanced which emphasizes how disparities are driven by officers trying to stay out of trouble given the competing demands faced by the ecological and organizational contexts in which they are embedded. In the case of stop-and-frisk, it is argued that neighborhood crime effects should largely drive disparities but with a disconnect from individual criminality as a result. Second, prior literature has relied on models that offer indirect and imprecise answers as to the substantive importance of neighborhood crime in driving disparities. A design is advanced to confront these issues, and findings reveal that neighborhood violence was overwhelmingly important in explaining racial disparities in stop-and-frisk, particularly because of the large volume of innocent minority individuals stopped in high-crime areas.