Type of Homicide and Variation in Regional Rates

A reanalysis of regional differences in homicide rates is conducted for two types of homicide, derived from the victim-offender relationship. A social — structural as opposed to subcultural, explanation is supported, but is found to have limitations. Social-structural variables, particularly poverty, are shown to be important predictors of differences in primary homicide rates (generally involving family members or friends), but are less important in explaining variations in non-primary rates (those generally involving strangers).