NCJ Number
195844
Journal
Justice Quarterly Volume: 19 Issue: 2 Dated: June 2002 Pages: 377-398
Date Published
June 2002
Length
22 pages
Annotation
This study used data on robberies and assaults from the 1992-99 National Crime Victimization Survey (NCVS) to determine whether victim, offender, and situational attributes moderated the deterrent effect of defensive gun use (DGU).
Abstract
Because the NCVS uncovers far fewer DGU's than virtually all other self-report surveys of gun use, critics have claimed that it contains an extremely high number of false-negatives; i.e., NCVS respondents fail to report their DGU's to interviewers for fear of exposing illegal behaviors or out of uncertainty that their actions were legally justified. Although the NCVS may undercount DGU's, other survey instruments are likely to overestimate their frequency. The dependent variable examined in the study was whether victims sustained injuries during the course of their victimizations. The independent variables represented victim, offender, and situational attributes, and each was measured as a binary variable. The key independent variable was whether or not an incident involved a victim who either used or threatened to use a gun in self-defense. Multinomial logistic regression analyses indicated that the victim's gender, household income, and location of the incident moderated the effect of DGU on injury. These findings suggest that although DGU provides a deterrent benefit for some in reducing victimization injuries, DGU does not significantly reduce the odds of injury for women or victims who reside in low-income homes. Findings also indicate that the deterrent effect of DGU is limited to urban settings. Thus, although using a gun will likely alter a situation, some incident-specific characteristics nonetheless apparently have a significant impact on the outcomes of armed resistance to crime. 5 tables and 27 references