Association for Public Policy Analysis & Management

Research on residential choice and neighborhood mobility suggests that neighborhood selection and assessment processes are strongly shaped by perceptions of racial composition and race-based preferences (Farley et al. 1978, Krysan 2002). At the same time, studies on perceptions of crime and disorder find that respondents’ beliefs about neighborhood safety are highly contingent on demographic composition and racial concentration, such that neighborhoods with larger minority populations are perceived to have higher levels of crime and disorder regardless of true crime levels (Quillian and Pager 2001). Given the common influence of neighborhood composition on both residential choice and perceptions of safety, one can imagine there exists a dynamic feedback loop through which perceptions of safety shape and are shaped by changing neighborhood demographics. Though Skogan (1986) and Hipp (2010) have both theorized that fear of crime may be an important mechanism driving the dramatic transformation of an area’s racial or ethnic and economic composition, no empirical studies to date have connected perceptions of neighborhood safety to neighborhood dynamics. As a result, sociologists have under-theorized the importance of perception and reputation as key mechanisms driving neighborhood evolution, even as governments, non-profits, and developers devote large amounts of capital to neighborhood redevelopment, branding, and marketing efforts.

To correct for this, this paper examines the relationship between perceptions and neighborhood dynamics, with a focus on perceptions of neighborhood safety. Focusing on 65 neighborhoods in Los Angeles between 2000 and 2008, I use longitudinal data from the Los Angeles Family and Neighborhood Survey, U.S. Census data, and Los Angeles crime data to examine the influence of crime levels and demographic trends on perceptions of neighborhood safety, including how changes in neighborhood demographic levels and crime rates affect retrospective recollections of neighborhood safety. Building on Quillian and Pager (2001) and Stinchcombe et al. (1980), I hypothesize that changes in crime rate have no effect on shifting perceptions of safety. Instead, I hypothesize that respondents perceive their neighborhoods to become less safe as the percent minority population in a neighborhood grows and that this relationship holds in retrospective models that consider how shifts in local demographics and crime levels change how safe one remembers their neighborhood to have been.

Initial descriptive evidence finds that perceptions of neighborhood safety are highly volatile, with a majority of respondents moving between categories of perceived safety across survey waves. Preliminary results from logistic regressions suggest that respondents’ heightened perceptions of neighborhood danger do not correspond to local crime rates. Additionally, I find some evidence that perceptions of neighborhood safety shift in relation to neighborhood demographic changes, particularly that increases in the percent black or Latino population within a neighborhood significantly predicts an increase in the odds of recalling ones’ neighborhood at an earlier time as safer than at present. Further research will extend these models to explore if disaggregating crimes by type and considering the spillover effects of nearby crimes or neighborhood composition influence respondents’ perceptions of neighborhood safety over time.