Firearms and Homicide: the Influence of the Weapon Substitution Hypothesis on the American Gun Control Debate

Abstract

For several decades now the gun control literature in the United States has continued to produce conflicting accounts in regards to the availability of firearms on the U.S's high rate of homicide. This thesis proposes that this conflict is, in part, due to the implicit and continued influence of Wolfgang's (1958) 'weapon substitution hypothesis'. Wolfgang's hypothesis proposes that the intentions of an assailant, whether they be to kill or injure, determined the weapon selected. Since guns are recognised as being highly lethal, all assailants who use such weapons were believed by Wolfgang to have been highly determined to kill. Among other negative effects, it is argued that Wolfgang's hypothesis introduced a mind-set to this controversial research area that has continued to influence the opinions of academics from both sides of the debate. This mind-set revolves around the consensually held belief that if a firearm assailant is believed to have been determined to kill then they would have been capable of killing in the absence of firearms. Importantly, this belief implies that the best possible predictor of lethal weapon substitution is if a firearm assailant is determined to kill. This is unlikely to be true. Mischel (1968: 135) has argued: 'A person's relevant past behaviours tend to be the best predictors of his future behaviour in similar situations.' After adapting Mischel's logic to fit the weapon substitution debate, the following predictor was produced. The best possible predictor of lethal weapon substitution to non-firearm weapons is whether people who had killed with firearms were as experienced at killing victims with non-firearm weapons as assailants who had actually killed with such weapons. This predictor was further developed into a more workable methodology that was capable of testing the validity of both Wolfgang's hypothesis and the consensually held belief it initiated. This methodology involved a comparison of the previous serious to fatal violent non-firearm convictions between those most likely to be determined firearm and knife killers. It was discovered that only 2.94 percent of those most likely to be determined firearm assailants and 25.23 percent of those most likely to be determined knife assailants had previous convictions for serious to fatal non-firearm assaults. This result was statistically significant to the p< 0.005 (Z score=2.84). After eliminating all other possible explanations for these results it was concluded that, in conflict with both Wolfgang's hypothesis and the consensually held belief, not all determined firearm assailants are likely to be capable of lethal weapon substitution. Furthermore, if some proportion of determined firearm assailants are unlikely to be capable of lethal weapon substitution, then those not so determined are likely to be even less capable. Therefore, it was concluded that inhibiting all potential firearm assailants from accessing guns would be likely to reduce the overall rate of homicide. However, this thesis was limited in being able to apply this conclusion to the United States because it was based on a New Zealand population. Nevertheless, it is argued that the perpetuation of the consensually held belief has inhibited the best possible predictor of lethal weapon substitution from being applied to a research area where prediction is of paramount importance. When the best possible predictor of lethal weapon substitution has not previously been applied, it therefore becomes more understandable why this research area is plagued by such controversy.