CATHERINE F. SPROULE AND DEBORAH J. KENNETT

TRENT UNIVERSITY PETERBOROUGH, ONTARIO

Nous avons comparé l'incidence des homicides par pistolets, autres armes à feu et autres moyens entre les états-Unis et le Canada de 1977 à 1983. Étant donné que le contrôle des fusils est plus rigoureux et plus étendu au Canada qu'aux états-Unis, surtout en ce qui concerne les pistolets, nous avons prédit que le taux des homicide par fusils serait plus faible au Canada qu'aux ÉtatsUnis et que la différence entre les taux des deux pays serait plus grande dans le cas des pistolets que dans celui des autres armes à feu. Nos constatations ont appuyé nos prédictions. En outre, nos constatations n'ont pas appuyé la thèse selon laquelle le contrôle des fusils est susceptible d'inciter les meurtriers à recourir à d'autres moyens que les armes à feu pour tuer.

We compared the incidence of killings by handguns, firearms other than handguns, and nonshooting methods between the United States and Canada for the years 1977 to 1983. Given gun control is more rigorous and pervasive in Canada than the USA, particularly for handguns, we predicted that the rate of killings by guns would be lower in Canada than in the USA and that the difference between the two countries' killing rates would be greater for handgun killings than for those committed by other firearms. Our findings supported the predictions. In addition, our findings did not support the claim that gun control may encourage murderers to use other methods than firearms to kill.

Several studies have not yielded clear or unequivocal evidence for the effectiveness of gun control in the United States (e.g.,⁷, ⁸, ¹⁰, ¹¹). On the other hand, Sproule and Kennett¹³ demonstrated a significant decrease in shooting homicide rates in Canada after the introduction of stringent gun control legislation in 1976. We propose that the discrepant findings regarding the benefits of gun control between Canada and the USA are likely the result of differences in the rigour and pervasiveness of gun control.

In comparison to Canada, gun control in the United States, particularly pertaining to handguns, is remarkably lax. There are an estimated 70 million handguns in the USA,¹⁴ but in Canada, handguns are restricted weapons. Permits for ownership of handguns in Canada may be obtained only by a) police and others, such as security personnel, who demonstrate a need for handguns in their work, b) members of bonafide gun clubs, c) bonafide gun collectors, and d) persons who demonstrate a need for handguns for protection.⁵ As far as we can ascertain, permits for handgun ownership have actually been granted only under the first three criteria. Further, infractions of the gun control provisions are indictable offences under the Criminal Code of Canada. In contrast, gun control provisions and criminal law in the United States are not federal but reside in each state or jurisdiction. In the USA, then, avoidance of local gun control regulations may be accomplished simply by crossing jurisdictional boundaries. Thus, not only are provisions for gun control much looser in most, if not all, of the American jurisdictions than in Canada, but also these less stringent regulations cannot be as well enforced as in Canada with its stricter federal gun control legislation.

The purpose of the present study is to examine the incidence of killings with guns in the two countries. Given the demonstrated effectiveness of Canadian gun control,¹³ it is expected that the rate of killing by guns will be lower in Canada than in the USA. Further, since the most rigorous regulations on firearms in Canada are for restricted weapons: i.e., handguns, the difference between the two countries' killing rates is predicted to be greater for handgun killings than for those committed by other firearms.

Canadian data for the rate at which victims of homicide were killed by handguns, by firearms other than handguns, or by methods not involving firearms, for the years 1977 to 1983 inclusive, were derived from Statistics Canada's annual publication on homicide.¹, ², ⁶ American data for the same years for victims dying by the same methods were derived from the 107th Edition of the Statistical Abstracts of the United States 1987.¹⁵

It is important to note that the American data are murder rates whereas the Canada data are for the more inclusive category of homicide (i.e., first and second degree murder, manslaughter and infanticide). Comparison of this more inclusive category of homicide to that of the more restricted category of murder should mitigate finding support for the hypotheses.

Equally important to observe is that standardized rates adjusted for age could not be calculated because the age of the victim was unknown. Although changes in age structure have been demonstrated to affect crude rates,⁴, ⁹ comparison of killing rates by different methods between two nations with similar demographics is unlikely to be affected by the use of crude rates.

The crude killing rates per 100,000, averaged over the years 1977 to 1983, are shown in Table 1 for killings by handguns, firearms other than handguns, and nonshooting methods for both Canada and the United States.

A 2 x 3 analysis of variance was performed on killing rates per 100,000 for the independent variables nation (USA vs Canada) and method of killing (handguns, firearms other than handguns, and nonshooting methods) over the years 1977-1983. Both main effects were significant. For all methods, the average American murder rate was significantly

higher than the average Canadian homicide rate [F(1,36) = 564.67, p <.0001]. As well, Scheffe's post hoc multiple comparison test for the significant main effect of method of killing [F(2,36) = 170.22, p <.0001] revealed that the mean killing rate for the two nations combined was significantly greater for nonshooting methods (2.55) than for handguns (2.16) [t(36,3) = 4.90, p <.05] and for firearms other than handguns (0.99) [t(36,3)= 19.60, p <.001]. In addition, the mean rate of killing by handguns (2.16) was significantly greater than the mean rate of killing by firearms other than handguns (0.99) [t(36,3) = 14.70, p <.001].

The significant main effect of method of killing, however, becomes less important in light of the significant interaction between nation X method of killing [F(2,36) = 423.39, p <.0001]. Employing Scheffe's post hoc multiple comparison test, all simple mean comparisons were significant except that no significant difference was found between the average Canadian homicide rate for handguns (0.28) and the average Canadian homicide rate for firearms other than handguns (0.67) [t(36,6) = 3.46, p> .05].

In addition, our prediction that the difference between the two countries' average killing rates over the years 1977 to 1983 would be greater for handgun killings than for those committed by other firearms was supported [t(12) = 15.98, p <.0001]. The mean difference in rates between the two countries for the years 1977-1983 for killings by handguns and for killings by firearms excluding handguns were observed to be 3.77 and 0.66, respectively.

Equally important are the following significant comparisons. First, the average USA murder rate for handguns (4.05) was found to be significantly greater than the average Canadian homicide rate for all methods of killing (2.73) [t(12) = 7.81, p <.001]. Second, even though the average American nonshooting rate of 3.31 is significantly higher than the average Canadian nonshooting rate of 1.79 [t(36,6) = 13.43, p <.001], significantly more Canadians were killed by nonshooting methods (1.79) than by shooting methods (0.94) [t(12) = 14.61, p <.001]. Third, the American shooting rate of 5.37 was observed to be significantly higher than the Canadian shooting rate of 0.94 [t(12) = 23.26, p <.001].

In strong support of our hypotheses and in a clear demonstration of the benefits of Canadian gun control, we found that Canadians kill less with firearms than Americans and that the difference between the two countries is larger for handgun killings (which are restricted weapons in Canada) than for those committed by other firearms. As well, American murder rates are higher for handgun killings than for killings by other firearms and for killings by nonshooting methods. Indeed, American murder rates for handguns are higher than the total Canadian homicide rate. The present findings combined with our earlier demonstration of a decrease in the Canadian shooting homicide rate after the introduction of gun control in Canada¹³ emphatically show that Canadian gun control, especially the provisions pertaining to handguns, does have the beneficial effect of saving lives.

In addition, these findings undermine the apparent claim of gun control opponents in their slogan 'people kill, guns don't,' which appears to mean that gun control does not affect the likelihood of killing but rather only the means by which death is accomplished. Also, the present findings allay a concern we raised regarding a finding in our comparison of methods of killing in Canada before and after gun control. Our results reported in a previous study¹³ indicated that there was a marginal tendency for nonshooting homicides to increase subsequent to gun control implementation, sugggesting that gun control may encourage murderers to use other methods than firearms to kill. The findings we present here, however, do not support the claim that gun control operates simply to induce killers to find alternate means to kill. More specifically, in Canada, the definition of handguns as restricted weapons is not associated with higher rates of killing with other more accessible firearms: the Canadian homicide rates for killing with handguns and with other firearms did not differ significantly. As well, the Canadian rate for other firearms was lower than the American rate for firearms other than handguns. Similarly, evidence against the claim that gun control facilitates the use of other methods was found for nonshooting killings. Although Canadians kill more with nonshooting than shooting methods, the Canadian nonshooting rate is significantly lower than the American nonshooting rate. Finally, in our earlier study¹³ we demonstrated that Canadian killers

using firearms killed more victims than Canadian killers using nonshooting methods: a killer with a gun is more likely to have multiple victims than a killer without a gun. The high American murder rate for shooting methods may be attributable, in part, to the multiple victims of a killer with a gun.

Additional support for the effectiveness of Canadian gun control comes from Sloan and his colleagues'¹² comparison of crime, assault and homicide rates in Seattle, Washington, and Vancouver, British Columbia, from 1980 through 1986. Although crime rates were similar in the two cities, assaults involving firearms were seven times higher in Seattle than in Vancouver and differences in homicide rates in the two cities were virtually all accounted for by the 4.8-fold greater risk of being murdered with a handgun in Seattle than in Vancouver.

Because Canadian gun control is clearly beneficial, we are concerned about recent reports of a proliferation in gun clubs and gun club memberships whose apparent aim is to obtain handgun permits and also of an increase in the number of handguns confiscated by Canadian customs at the border.³ Because Canada's favourable situation regarding murder relative to the United States is to a large measure the result of Canadian gun control, Canadians must be vigilant against any erosion of our gun control provisions. Further, because gun ownership (at least in Detroit) is inversely related to individuals' confidence in collective institutions to protect their security of person and property,¹⁶ maintenance of Canadians' confidence in the police and justice system to protect our collective security is an important means by which to deter gun acquisition. Above all, the important role played by gun control in Canadians' collective security must be stressed and more widely recognized.

1. Canadian Centre for Justice Statistics, Law Enforcement Statistics Section. Homicide in Canada 1983: A Statistical Perspective. Ottawa: Statistics Canada. Catalogue 85-209.

2. Canadian Centre for Justice Statistics, Law Enforcement Statistics Section. Homide in Canada 1982: A Statistical Perspective. Ottawa: Statistics Canada. Catalogue 85-209.

3. Canadian Press. "More Americans having handguns seized at border, statistics show." Toronto Star. Toronto, February 1, 1988 .

4. Conklin, G. H. and M. E. Simpson, "A demographic approach to the cross-sectional study of homicide." Comparative Social Research. 1985 , 8171-185.

5. Criminal Law Amendment Act, 1977 .

6. Justice Statistics Division. Homicide Statistics 1977. Ottawa: Statistics Canada, Catalogue 85-209.

7. Kleck, Gary. "Capital punishment, gun ownership and homicide." American Journal of Sociology. 1979 , 84, 882-910.

8. Kleck, Gary. "Evidence that 'Saturday Night Specials' are not very important for crime." Sociology and Social Research. 1986 , 70, 303-307.

9. Lee, G. Won. "Are crime rates increasing? A study of the impact of demographic shifts on crime rates in Canada." Canadian Journal of Criminology. 1984 , 26, 29-41.

10. Lester, D. "The murder of police officers in American cities." Criminal Justice and Behavior. 1984 , 11, 101-113.

11. Lester, D. and M. E. Murrell. "The influence of gun control laws on personal violence." Journal of Community Psychology. 1986 , 14, 315-318.

12. Sloan, J. H., A. L. Kellermann, D. I. Reay, J. A. Fenis, T. Koepsell, F. P. Rivara, C. Ric, L. Gray , and J. Logerfo. "Handgun regulations, crime, assaults, and homicide: A tale of two cities." New England Journal of Medicine. 1988 , 319, 1256-1262.

13. Sproule, C. F. and D. J. Kennett. "The use of firearms in Canadian homicides 1972-1982: The need for gun control." Canadian Journal of Criminology. 1988 , 30, 31-37.

14. Tenszen, Michael. "Keep handgun ban, U.S. police official warns Canadians." Globe and Mail. Toronto, August 27, 1986 .

15. United States Department of Commerce. Statistical Abstract of the United States 1987. 107th Edition.

16. Young, R. L., D. McDowall, and D. Loftin. "Collective Security and the ownership of firearms for protection." Criminology. 1987 , 25, 47-62.