Homicide and medical science: is there a relationship?

• Author: Andresen, Martin A.

1. Introduction

Harris, Thomas, Fisher, and Hirsch (2002) investigate the relationship between homicide, aggravated assault, and medical science for the United States as a whole and call the disparity between the trend in the rate of aggravated assault, on the one hand, and that in the rate of homicide, on the other, a "paradox" in need of resolution (129). (See Figure 1 for the homicide and aggravated assault rates in the United States, 1960-2000.) They claim that advances in medical science decrease trauma-induced lethality and, therefore, homicide. Violent crimes that would have been homicides in the past are now aggravated assaults because such crimes do not have the "outcome of the victim's death" (2002: 128). More specifically, Harris et al. (2002) cite the strong relationship between the lethality of criminal assault (measured as the number of homicides relative to the number of homicides plus aggravated assaults), on the one hand, and variables representing access to medical care, on the other--the presence of a hospital (access to medical resources), the presence of a trauma centre (access to the latest in medical science), and the presence of a regional trauma system (a medical science network). After presenting evidence of this relationship, they argue that research on the causes and prevention of homicide is better focused on serious assaults, of which only a small percentage result in a death--the lethality approach.

[FIGURE 1 OMITTED]

Though an impressive finding, with strong implications for criminal justice and health policy, this study is not without its limitations: in its current form, the lethality approach is coarse, and the index of lethality is quite broad, possibly capturing too many criminal assaults that cannot be considered potential homicides. As well, although Harris et al.'s (2002) analysis was carried out across time and regions within the United States, an analysis of another nation with similar medical technology is needed to discover whether the relationship within the United States is generalizable; finally, the "fact," so styled, of the disparity between trends in the rates of aggravated assault and of homicide--the point of departure for their study--is on shaky foundations.

This article contests the results of Harris et al. (2002) by critiquing their definition of lethality, by examining their findings in light of a cross-country (Canada-United States) comparison, and by calling into question the existence of the disparity they are trying to explain. The positive impacts of having advanced medical resources are not contested here, only the premises from which Harris et al. (2002) launch their investigation. The next section (section 2) lays the foundation for a comparison of Canadian and U.S. homicide/lethality rates and presents the data and methodology employed to contest the Harris et al. (2002) results. The medical science hypothesis is assessed in section 3. Section 4 concludes that the medical science hypothesis is not robust.

2. Data and methodology

   2.1. Can Canadian and U.S. homicides be compared?

   In order to be confident in accepting or rejecting the medical science hypothesis in light of a cross-country comparison, it is important that the modes of homicide in Canada and the United States be similar. The modes of homicide in each country do not need to be exactly the same, but they must be similar enough to be sure that cross-country differences in the lethality of criminal assault are not due to differences in the nature of those assaults. Most importantly, the modes of homicide within each country must be relatively constant over the study period. Otherwise, a change in lethality may simply reflect a change in the modes of homicide.

   As shown in Figure 2, firearms have been the most frequent mode of homicide in the United States: over the past 25 years, 60 to 70% of all homicides were committed with firearms--45 to 55% with handguns alone. Importantly for the comparison, "penetrating-wounds homicides" (homicides using guns or knives) in the United States--the type of homicide relevant for considering advances in medical technology--were relatively constant over the entire study period. There was a noticeable rise, in the 1990s, in the use of handguns for homicide, with a corresponding fall in the use of knives (Fox and Zawitz 2002).

   [FIGURE 2 OMITTED]

   Though slightly lower than in the United States (see Figure 2), the proportion of homicides committed in Canada with firearms or knives has also been relatively constant, ranging between 57 and 65% over the past 25 years. There was a decrease in firearms-related homicide in Canada over the study period, but knife-related homicide rose correspondingly. Therefore, the proportion of penetrating-wound homicides in Canada was relatively constant. This is critical for a proper assessment of the Harris et al. (2002) premises because the proportion of penetrating-wound homicides in the United States has also been constant. As in the United States, the use of handguns in Canadian homicide has risen dramatically over the past 15 years: the proportion of handguns in relation to all methods for committing homicide has risen from an average of 10% to an average of 20%, and the proportion of handguns in relation to all firearms as a method for committing homicide has risen from an average of 29% to an average of 64%. Therefore, handgun use in Canadian homicides has increased dramatically in absolute terms and even more so as a proportion of homicides committed using firearms. One could argue that the use of rifles and shotguns is more prevalent in Canada than in the United States, given the availability of handguns in the United States. However, it should be clear from the data presented here that handgun use in Canadian homicide is high and is on the rise. (2) I do not argue that the wounds inflicted with different weapons are qualitatively similar, but the Harris et al. (2002) argument is based on the treatment of penetrating-wound trauma, broadly defined.

   Because the percentage of penetrating-wound homicides in Canada remained constant over the study period, it is expected that Canada's lethality will be found to behave in a manner similar to that of lethality in the United States (downward trend over the entire study period) if the medical science hypothesis is correct. Of course, there are scores of other determinants of homicide, but the analysis at hand is concerned with the lethality of those homicides, not with what caused them to occur. And given that the modes of homicide (proportions of penetrating-wound homicides) have remained relatively constant over the study period, I am confident that any changes in the Canadian lethality index are a result of advances in medical science, just as Harris et al. (2002) argue was the case in the United States.

   One final issue regarding the comparability of Canada and the United States is the question of medical resources. There are two issues at stake: the availability of trauma care and the quality of that trauma care. Both the United States and Canada have experienced significant growth in the...