Spatial Choice of Auto Thefts in an Urban Environment

• NCJ Number: 215782
• Journal: Security Journal Volume: 19 Issue: 3 Dated: July 2006 Pages: 143-166
• Author(s): Yongmei Lu
• Date Published: July 2006
• Length: 24 pages
• Publication Format: Article
• Publication Type: Report (Study/Research)
• Language: English
• Country: United States of America

Annotation

This article reports on analyses of auto thefts in Buffalo, NY, using geographic information systems (GIS) and exploratory spatial data analysis.

Abstract

Auto thefts were found to be clustered in the city. Some areas of the city had more auto thefts because there were more vehicles there; other areas had high concentrations of auto thefts because certain characteristics of these areas put the vehicles at a higher risk of being stolen. The research also showed spatial patterns of auto thefts as a linear distribution along urban streets, which suggests the investigation of the impact of street networks on auto-theft patterns. The major roads and the roads directly connected to the major roads in Buffalo were sites of most of the auto thefts. Along these streets, there were linear "hot spots" of auto thefts. Apparently certain parts of the streets were more attractive to offenders than other parts. This may be related to accessibility and patterns of socioeconomic activities along the streets. The analyses showed that apartment complexes, downtown row housing (detached), three-family year-round residences, and commercial parking lots were at higher risk as sites targeted by auto thieves. These findings provide strong support for various crime-event theories, including routine activity theory and rational choice theory. These findings suggest the targeting of identified "hot spots" for various crime prevention measures. The addresses of the auto theft locations were obtained from crime reports and decoded into a standard street address format. Geocoding was conducted to map these addresses onto a city street map. The location map of auto thefts was the basis for further analyses. Census 2000 data were used to extract maps of census blocks and block groups in the study area. The census attribute of "vehicles available" was obtained as an indicator for target size. 7 tables, 4 figures, and 49 references