Date of Award

Winter 1990

Project Type


Program or Major


Degree Name

Doctor of Philosophy

First Advisor

Murray A Straus


Little homicide research has been done which is both race, and victim/offender relationship specific. This dissertation provides such an analysis, with a quantitative description and analysis of murders which occurred between black family members in the United States between 1976 and 1987.

Data for this project was obtained from the FBI's Supplemental Homicide Report. The first half of this dissertation examines patterns (and time trends) of family, acquaintance, stranger, and spouse, parent, and child homicide rates for blacks in America in terms of variables such as age, gender, weapons and circumstances.

Many of the most interesting descriptive patterns discovered related to the gender variable. For example, black women were found overall to have relatively larger levels of involvement in family homicide events (relative to the involvement of men) than in other relationship categories; black women were less likely to be victims of spousal homicide than were black men (with other research finding the opposite gender pattern holding for whites). In addition, examination of time trends found that family homicides among blacks show greater levels of change, i.e., they are less static, than are acquaintance or stranger murders.

The second half of the dissertation involved the construction of a regression model, predicting variation in black family and non-family homicide rates across a sample of 86 American cities. While many of the same variables were useful in predicting both relationship types of homicide (e.g., measures of income deprivation) there were some notable differences. For example, a Southern regional variable had no effect on the rate of non-family homicide, but was negatively related to the incidence of family homicide among blacks. The percent of black children living with two parents was a positive predictor of family homicide, and a negative predictor of non-family homicide. Finally, the non-family homicide rate accounted for only an additional 8% of the variance in the rate of family homicide when other factors were controlled for.