This study explores the relationship between neighborhood's social capital and burglary victimization by using mail survey data conducted in the urban area of Japan. We examined how crime control effects of social capital vary, depending on changing of geographical range of the “neighborhood” used in the analyses, from 60 to 500 meters in 10 meter increments. We summarize the results as follows: (1) for trust, norms of reciprocity and the principal component score of the social capital index, the larger geographical ranges of “neighborhood” used in analyses are, the stronger their crime control effects are; (2) social participation doesn't have significant crime control effects; (3) trust shows strong crime control effects even in narrow geographical ranges. These findings suggest that there may be multiple crime control mechanisms through social capital in a certain geographical range of the neighborhood.
