RACE AND VIOLENT CRIME IN THE SUBURBS

It is well known that crime rates, when examined by residential area, are positively correlated with racial composition. This is usually int erpreted to mean that racial composition affects crime rates, although there is debate over why We consider an alternative interpretation: t hat reciprocal causal effects exist between these two variables. The c rime rate itself may change the racial composition of an area by makin g it a less desirable place in which to live and invest. This hypothes is is tested with longitudinal data for a national sample of suburbs f or the period 1970 to 1990. We find significant and approximately equa l causal effects in both directions; specifically, it is the robbery c omponent of crime that affects racial composition. High robbery rates ape associated with black population growth while stimulating white fl ight.