Rates of homicide involving intimate partners have declined substantially over the past 25 years in the United States, while public awareness of and policy responses to domestic violence have grown. To what extent has the social response to domestic violence contributed to the decline in intimate-partner homicide? We evaluate the relationship between intimate-partner homicide and domestic violence prevention resources in 48 large cities between 1976 and 1996. Controlling for other influences, several types of prevention resources are linked to lower levels of intimate-partner homicide, which we interpret in terms of their capacity to effectively reduce victims' exposure to abusive or violent partners. Other resources, however, are related to higher levels of homicide, suggesting a retaliation effect when interventions stimulate increased aggression without adequately reducing exposure. In light of other research on deficiencies in accessing and implementing prevention resources, our results suggest that too little exposure reduction in severely violent relationships may be worse than none at all.
In the United States, rates of homicide involving "intimate partners"-spouses, ex-spouses, boyfriends, girlfriends-have declined substantially over the past 25 years. Public awareness of and policy responses to domestic violence have increased during the same period. The coincidence of the two trends leads naturally to the question of their relationship: To what extent has the social response to domestic violence contributed to the decline in intimate-partner homicide? Research evidence addressing that question is highly limited, but the few existing studies suggest that domestic violence resources and policies such as hotlines, shelters, and legal advocacy programs may be associated with lower rates of intimate-partner homicide, net of other influences (Browne & Williams 1989; Dugan, Nagin, & Rosenfeld 1999).1
In this article, we address the relationship between intimate-partner homicide and domestic violence resources for a larger number of places over a longer period of time and with a considerably richer set of outcome and resource measures than used in previous research. Building on the research by Dugan, Nagin, and Rosenfeld (1999), we interpret that relationship in terms of the exposure-reducing potential of domestic violence resources. Simply put, those policies, programs, and services that effectively reduce contact between intimate partners reduce the opportunity for abuse and violence. However, we also assess the alternative possibility that, under certain conditions, domestic violence resources provoke a retaliation effect. Such an effect might occur, for example, if a protection order or other legal intervention directed at an abusive partner increased the level of stress or conflict in the relationship without effectively reducing victim exposure. We evaluate the exposure-reducing and retaliation effects of a broad range of domestic violence resources on levels of heterosexual intimate-partner homicide by victim sex, race, and marital relationship to the offender for 48 large U.S. cities between 1976 and 1996. Further, because we anticipate that other factors can affect the exposure between violent intimates, we control for changes in marriage and divorce rates, women's status, and other time- and place-varying influences.
The growth in domestic violence resources in the United States occurred during a period of declining intimate-partner homicide. The coincidence of the contrasting trends in intimate-partner homicide and social response is especially notable because the overall rate of homicide is trendless during the same period.2 The general decline in intimate-partner homicide varies substantially by victim sex, race, and marital relationship to the offender. Larger decreases have occurred for males, African Americans, and married victims (including ex-spouses) than for females, whites, and unmarried intimates (Greenfield et al. …