How Context Affects Politics: Essays on Causality and Measurement

Abstract

How do contextual features of the neighborhoods where we live, work, and play influence civic and political behavior? Micro-level forces, often overlooked in the study of broad social and political processes, shape our decision-making in myriad ways. I propose that who and what we encounter as we go about our everyday lives affects the policies we support, as well as how we vote and interact with government. This dissertation explores the effects of social and physical context through a series of empirical studies, beginning with the most micro of settings --- individuals on a city street --- and progressing to the more macro --- voting precincts, neighborhoods, and municipalities. The first study describes a placebo-controlled field experiment in which encounters between the well-off and the poor in affluent neighborhoods decrease the willingness of the former to support redistributive public policies. The second study uses geo-located pedestrian traffic counts from a network of live video camera feeds, overlaid with municipal hotline (311) data and crime incident reports, to demonstrate how the degree to which city-dwellers function as "eyes on the street" is highly context-dependent. The third study reveals how violent protests can be mobilizing events that shift local voter support in favor of public goods associated with the rioting group, and that this mobilization has enduring downstream effects. The final study leverages the end of legal, state-sponsored segregation in South Africa, along with natural physical geography, to show that sustained racial isolation increases white South Africans' likelihood of voting against historically black African or other non-white parties.