UC Irvine

This study attempts to disentangle the extent to which residents are systematically biased when reporting on the level of crime or disorder in their neighborhood.  By utilizing a unique sample of households nested in household clusters, this study teases out the degree of systematic bias on the part of respondents when perceiving crime and disorder.  The findings are generally consistent with theoretical expectations of which types of residents will perceive more crime or disorder, and contrast with the generally mixed results of prior studies that utilize an inappropriate aggregate unit when assuming that residents live in the same social context of crime or disorder.  Estimating ancillary models on a sample of respondents nested in tracts produces mixed results that mirror the existing literature.  I find that whites consistently perceive more crime or disorder than their neighbors.  I also find that females, those with children, and those with longer residence in the neighborhood perceive more crime or disorder than their neighbors.