Citation

Harries K. Geoforum 2006; 37(3): 404-416.

Copyright

(Copyright © 2006, Elsevier Publishing)

DOI

10.1016/j.geoforum.2005.09.004

PMID

unavailable

Abstract

A salient characteristic of the geography of crime in the US is the presence of extremely sharp geographic variations. These variations may be significant indicators of local environmental inequalities and may have implications for fear of crime and crime contagion, and may also be indicators of potential or actual neighborhood instability. Such micro-level variations are not generally apparent on small scale maps of urban crime. Previously, micro-level analysis of this gradient phenomenon has been inhibited by the confluence of large volumes of data over large areas, a plethora of possible boundaries that might be used for the purpose of data aggregation, and practical difficulties in the identification of gradients in the context of small units of analysis. The present study attempts to identify steep crime gradients and to characterize the physical and social circumstances under which they occur. Analysis was based on 97,880 geocodable incidents reported in 2000 in Baltimore County, Maryland. Crime densities were calculated for 5324 census blocks that experienced at least one crime incident. A steep gradient for the purpose of this research was the juxtaposition of blocks in the highest and lowest quintiles in terms of crime density. Using residential and commercial land uses as a filter, some 259 blocks satisfied the gradient criterion. Further analysis linked these blocks to their parent block groups for the purpose of identifying their social attributes. In addition, six clusters of blocks were investigated in the field. A typology of adjacencies identified six categories.