Research on Gun Deaths
Read the section and watch the video on the research concerning gun deaths. What are the general conclusions regarding the research? Does this make sense to you?
Reading
Gun Violence Research
To fill in the gap on gun violence research a number of philanthropic organizations funded organizations focused on the topic such as The National Collaborative on Gun Violence Research, The RAND Corporation, and the National Gun Violence Research Center. A small number of states have created research centers on university campuses including the University of California Firearm Violence Research Center, the Rutgers Center on Gun Violence Research, the Johns Hopkins Center for Gun Policy and Research, and the Harvard Injury Control Research Center (Van Brocklin, 2019). Together, these organizations and centers have conducted much of the research discussed below.
Guns for Self-Defense
One of the justifications that the Supreme Court gave for granting an individual right to handgun ownership in Heller was for self-defense. If criminals have a reasonable fear that homeowners are armed, they should be less likely to commit violent home invasions or burglaries. Thus, homicide rates might be lower in communities with high levels of gun ownership. On the other hand, higher levels of gun ownership could lead to higher homicide rates as guns might become more easily available to commit robberies and other violent crimes. Gun availability could also lead to an escalation to murder in domestic arguments.
The case-control research on this topic studies the benefits and risks associated with keeping a gun in the home. The Kellerman study mentioned earlier was conducted in three counties in Ohio, Tennessee, and Washington. They identified all homicides that occurred in a home involving residents age 13 or older over a five-year period. They obtained information on the homicides from newspaper stories, police interviews, obituaries, and calls to funeral homes. A control group was created that matched the victims on age, gender, race, and neighborhood. There were more handguns in the victim’s home than the controls. Gun ownership was statistically associated with a greater likelihood of being killed by a family member or intimate acquaintance. Only 4 percent of the victims were killed by strangers and there was forced entry in 14 percent of the cases (Kellermann, et al., 1993). An additional analysis of this data found an even stronger risk of homicide victimization for women by family members and intimates (Bailey, et al., 1997).
Wintemute and colleagues (1999) conducted a cohort study of 235,000 people who legally purchased a handgun from a licensed dealer in California. Firearm purchase was related to lower than expected odds of homicide victimization for men but significantly higher than expected odds for women. While these studies are suggestive there is not a causal link between gun ownership and homicides. For example, we do not know whether the gun purchased was used in the homicide. We do not know whether someone felt at risk of victimization, bought a gun, and was subsequently killed.
Several ecological studies have been conducted that compare similar geographic areas in firearm availability and homicide. Killias (1993) conducted a study comparing 14 developed countries including 11 in Europe and the U.S., Canada, and Australia. The data on gun ownership came from 28,000 respondents in the International Crime Survey of 1989. There was tremendous variation in gun ownership with just 2 percent in the Netherlands and 48 percent in the U.S. There was a statistically significant relationship between firearm ownership and homicide rates. There was no statistical relationship between ownership and non-firearm homicide.
Hemenway and Miller (2000) conducted a similar study using the highest income countries as defined by the World Bank (26 countries). They used Cooke’s Index, the average of the percent of gun homicides and the percent of suicides with a gun as proxy measures. Both measures were strongly related to gun ownership. When the U.S. was removed as an outlier, the association remained. When the countries were weighted to control for differences in population, the relationship between gun ownership and homicide was even stronger.
One study compared Seattle, Washington and Vancouver, Canada. These two cities are close together and are very similar in geography, climate, history, and socioeconomic status of residents. The primary difference is that 41 percent of Seattle residents own a gun compared to 12 percent of Vancouver residents. The study compared homicides over six years. The risk of being a victim of homicide by firearm was five times greater in Seattle compared to Vancouver. The risk of homicide by other means was not significantly different between the two cities (Sloan, et al., 1988).
Several other international studies have come to similar results. In comparison to other high-income nations, the U.S. has eleven times the firearm homicide rate. The rate of firearm ownership and homicide is strong and statistically significant. The relationship is not as strong when the U.S. is removed from the analysis. Researchers believe this is because of the licensing, registration, and safe storage requirements of other nations (Hepburn & Hemenway, 2004).
A number of cross-sectional studies among regions, states, and cities in the U.S. have found that higher levels of guns is associated with higher levels of homicide overall, and especially homicide by guns. Comparisons between states is especially striking. After controlling for poverty, unemployment, crime rates, and urbanization, there is a strong relationship between gun prevalence and homicides among all age groups and both genders (Hepburn & Hemenway, 2004). Additionally, all of the types of studies in this section have been applied to suicide and have found very similar results. There is a strong relationship between gun ownership or gun prevalence and suicide (Stroebe, 2013). While all of these studies are strongly suggestive that having a gun in the home and gun prevalence in general is strongly related to homicide and suicide risk, there are data limitations in the sense that we have to use proxy measurements of gun ownership. The only national database on gun ownership maintained by the ATF is not available for research purposes.
