![Sample our Law journals, sign in here to start your FREE access for 14 days]({"type":"image" , "src" : "/sda/5942/TOC-Subject-Sample-2021-Law.jpg"})
Abstract
The dangerousness of domestic violence escalates when suspects use weapons against victims or responding officers. Nevertheless, only a few studies have examined the dynamics of weapon use in domestic violence. While supporting the situational approach, the limited literature and relevant theories suggest the need for weapons classification and multilevel research. Using over 9,400 domestic violence cases across 423 census tracts that were responded to by the Houston Police Department in 2005, hierarchical linear models examine the correlates of weapon use by suspects. Results indicate that situational and neighborhood factors are distinctively associated with each type of weapon. Implications for future research and policy are discussed.
Notes
1. A national survey report by Black et al. (2011) indicates that about 7 million women and 5.7 million men in the United States experienced rape, physical violence, or stalking in 2010. We divided the total number by 365 days.
2. CitationCampbell (1986) invented a danger assessment tool for domestic homicide, indicators of which include escalation of violence; use of weapons, alcohol, or drugs; and psychological or physical abuse, among others. CitationKane (1999) reviewed police records of domestic violence reports and examined the predictors of an arrest decision. Various types of weapon used against victims were considered risk factors and were found to significantly contribute to the officer's decision to arrest the suspect. Lastly, CitationRobinson (2006) introduced a risk assessment tool that South Wales Police in Cardiff adapted from Campbell's danger assessment tool.
3. Routine activities theory is also known as victimization theory. However, as CitationOsgood et al. (1996) noted, “yet the theory's basic prediction, that crime depends on routine activities, pertains to individual offending as well” (p. 636).
4. CitationRabe-Hemp and Schuck (2007) raised questions for previous studies where officer injuries were used as indicators of officer safety threats. They argued that the injury is only an “accidental” (p. 414) aspect of the overall danger; it does not reflect the totality of the safety threat. They thus measured the safety threat by suspect's weapon use against officers and profiled the correlates.
5. They also found that neither relationship status nor suspect's substance or alcohol use had an association with weapon use. Suspects with restraining orders were more likely to use a weapon.
6. This finding coincides with an intimate partner homicide study where females were more likely than males to use a knife (e.g., CitationSwatt & He, 2006).
7. Although not particularly about weapon use, a few studies have factored in neighborhood characteristics in domestic violence research (CitationVan Wyk, Benson, Fox, & DeMaris, 2003). Specifically, concentrated disadvantage and residential instability appear to exert a significant influence on the prevalence of domestic violence (CitationBenson, Fox, DeMaris, & Van Wyk, 2003; CitationMiles-Doan, 1998).
8. CitationKernsmith and Craun (2008) could not profile weapon use with regard to particular types of weapon due to insufficient caseloads for each type. Neither could CitationRabe-Hemp and Schuck (2007), because the details on the types of weapon used were not available.
9. A little more than half of the total domestic violence cases occurred in Houston in 2005 had no suspect present when police initially responded, as was the case in previous studies (CitationFeder, 1996; CitationFernandez-Lanier, Chard-Wierschem, & Hall, 2002). Those cases were considered invalid as we cannot infer any suspect demographic information.
10. Among the excluded group, 26% used other weapons, 3.4% used cutters, 2.1% used firearms, and 68.5% used body force. Among the included group, 26.0% used other weapons, 7.1% used cutters, 3.7% used firearms, and 63.2% used body force.
11. While the “other weapons” data comprised the largest category among the weapons, our data had no specification of the other weapon used. As a reference, CitationKernsmith and Craun's (2008) dataset had a specific weapon category of “phone,” which accounted for the largest portion of the weapons (36% among the weapons-used incidents). Further examination would be fruitful given that other weapons might cause more serious injuries than would firearms or knives (U.S. Department of Justice, 2003).
12. The -71 is not a coding error; one unusual case involved a 22-year-old male allegedly using a firearm (aggravated assault) to threaten a 93-year-old female complainant in their apartment. Responding officer(s) arrested the suspect.
13. Body force was set as a reference category not only for its large caseloads, but also for the sake of convenient interpretation of the statistical analysis results.
14. We tested whether the log-odds of weapon use (relative to body force use) varied between census tracts for each type of weapon. This unconditional model reported that there was a significant variation in the log-odds of cutter use (Var (u 2) = .202, p < .05) or other weapon use (Var (u 3) = .046, p < .01) relative to body force use across census tracts. By contrast, the log-odds of firearms use to body force use did not vary significantly (Var (u 1) = .253).