Communities and change: Racial and ethnic transition and crime in Los Angeles

This dissertation takes a multi-faceted look at the relationship between community change and crime. Social scientists have long argued that social and demographic factors play a role in the persistence of crime in particular communities. One of the most argued factors is that residential stability has a positive effect on crime rates; it is in communities where the population fluctuates that crime flourishes because the unstable population is unable to unite to address local incivilities and crime. Indeed, residential mobility is one of the three main criminogenic community factors of Shaw and McKay's seminal work on which most neighborhood effects literature is based. In this dissertation I examine a specific form of residential instability: racial and ethnic turnover. Beyond the theorized detrimental effect of residential turnover generally, racial and ethnic transition specifically can potentially fuel animosity and/or competition between groups and exacerbate the crime problem. Therefore, I focus on the dynamics of racial/ethnic change in the City of Los Angeles and examine (1) change in the city as a whole between 1990 and 2000; (2) African American to Latino transition in South Los Angeles and the effect on inter- and intra-group youth violence; and (3) the impact of school-specific and community racial and ethnic change on school serious offense rates. In the process, I acknowledge the importance of reciprocal causal effects and spatial effects in my analyses and employ methodological techniques to account for the complexities of community-based research.;Major findings include the robust effect of changes in the crime rate on neighborhood racial and ethnic composition. This holds true regardless of crime type, specification by victim---offender race/ethnicity, and whether racial and ethnic change is defined as heterogeneity or racial/ethnic churning (turnover). Surprisingly, there is limited evidence supporting social disorganization theory: change in the racial/ethnic composition of student enrollment significantly affects changes in offense rates in schools but not in neighborhoods. The characteristics of geographically proximate neighborhoods are significant predictors of change in communities and schools. Overall, this study demonstrates the importance of considering alternative definitions of residential instability, and reciprocal relationships and spatial effects in neighborhood-level research.