In Britain, Herbert (1979) and Baldwin and Bottoms (1976) found that spatial variations in offender rates are associated with variations in household tenure. Areas dominated by owner-occupied dwellings show lower rates than would be predicted by their socio-economic composition alone. The owner-occupied dwellings show the lowest crime rates. Public-sector housing display relatively high levels of criminal activity (Cater and Jones, 1989). Housing and criminality are related because social groups with a greater propensity to commit crime are concentrated in certain types of housing. Some areas have a large numbers of offenders residing within them because the type of housing in such areas is more available to those individuals at greater risk of offending. Examples include areas of privately rented accommodation with over-representations of young single males, and transient populations (Bottoms and Wiles, 1988). Individuals who rent a home may have a higher propensity to commit crime. Baldwin and Bottoms (1976) strongly suggested that the housing market might be relevant to the spatial distribution of offender rates over other variables, for example, social class distribution of households in the area. Wikström (1991) also found similar results, based on the path-model analysis for offender rates in different areas of Stockholm.