Perpustakaan Digital ITB

The lack of secure bicycle parking is a serious but often neglected issue that discourages bicycling. Classical environment criminology theories try to explain the pattern of bicycle theft but provide limited insights into the relationship between crime and the built environment. This study examines the association between built environment factors and bicycle theft using a zeroinflated negative binomial model to account for data over-dispersion and excess zeros. The assembled dataset provides variables pertaining to the road network, land use, bicycle travel demand, and socio-demographics. The key findings are as follows: (1) Bicycle theft is more likely to occur in areas for commercial purposes, areas with a high population or employment density, and areas with more bike lanes or sidewalks. (2) Bicycle theft is likely to occur at sites with more bike racks or bus stops. (3) Bicycle theft is more likely to occur at mid-blocks than at intersections. (4) Bicycle theft is more likely to occur in neighborhoods with a greater percentage of socially disadvantaged people and in neighborhoods where residents’ median age is lower. (5) The likelihood of losing a bicycle is lower in areas with more bicycle trips. In general, the number of bicycle thefts increases in dense areas with more targets and decreases with greater natural guardianship provided by more passersby. With respect to policy implications, governments and transport planners should implement a geographically-differentiated surveillance strategy, encourage bicycling, improve the visibility of bike racks to the public, and promote surveillance and natural guardianship in densely developed areas.