New estimates of the determinants of urban crime (original) (raw)

1975, The Annals of Regional Science

New tests of the utility maximizing model of criminal behavior, first proposed by Becker, are performed, These tests use newer data than has previously been available. The data are for 1960 and 1970 and for sixty cities. Ordinary least squares and two stage least squares methods are used. We find that probability of conviction has a significant deterrant effect, while length of sentence does not, thus indicating that society is spending enough on detecting crime to make only risk preferring individuals engage in criminal activity. We also find that percentage of non-whites is positive and significantly related to personal, but not to property, crime. This result, which differs from conclusions reached by others, leads us to propose that non-white personal crime may be a method of self-enforcement of property rights by blacks, necessitated by less police protection for this group. While the data does not allow a direct test of this hypothesis, it does seem consistent with our results.