Firearms, Accidental Deaths, Suicides and Violent Crime: An Updated Review of the Literature with Special Reference to the Canadian Situation

5. Firearms and Violent Crime

5. Firearms and Violent Crime

Violent crimes, including homicide, will never disappear entirely. Research identifies several factors to be associated with violence; the availability of firearms clearly being one of them. The extent to which violent crime can be prevented by reducing or controlling access to firearms is widely debated. Different types of violent crimes call for different prevention methods. The role of various factors associated with violent crime, including such situational determinants as the accessibility of firearms, is not necessarily the same in one kind of violent crime as in another.

5.1 Firearms and Homicide

Although instructive, trends in Canada’s homicide rate are sometimes difficult to interpret. For instance, there were fewer homicides in 1950 than in any year between 1926 and 1998. From 1950 to 1965, the rate rose gradually. Between 1966 and 1975, the rated increased by a dramatic 250 percent; from 1.2 to 3.0 per 100,000 population (Silverman and Kennedy, 1993: 34). Since then, there has been a fairly consistent decline in the homicide rate, from 3.0 in 1975 to 2.1 in 1996 (Hung, 1997). Research evidence does not explain why the homicide rate began to turn around in 1975; it may relate to social, demographic and other factors, not to mention the impact of new criminal justice practices to incapacitate offenders and deter recidivism.

Between 1961 and 1990, 40 percent of murderers killed their victims with a firearm (Silverman and Kennedy, 1993: 97). In 1996, the firearm homicide rate was 0.7 per 100,000 population. Based on police reports, 211 people were murdered with a firearm, representing one-third of the 633 homicides committed during that year. That rate is consistent with that of the previous 20 years, during which the proportion of homicides involving a firearm averaged about 32.9 percent. The percentage was higher before 1975 when it used to fluctuate between 40 and 48 percent. The proportion of homicides that involve a firearm varies regionally as well: in 1996, it was highest in Nova Scotia at 44 percent; it was 41 percent in Quebec; 38 percent in British Columbia; 18 percent in Manitoba; and it was lowest in Saskatchewan at 13 percent (Hung, 1997a).

5.2 Characteristics of Firearm Homicide

Firearms play different roles in homicides depending on the circumstances behind the incident; what type of incident it is; the age and sex of both the victim and the offender; the relationship between them, and other factors. For instance, a distinction is often made between a primary homicide, in which the offender intends to cause serious injury or death to the victim, and a secondary homicide, which happens when another crime is being committed (Goetting, 1995). Primary homicides are more common. They are usually directed at an acquaintance, and often one with whom the offender has had an intimate relationship.

Most murderers are male. According to the Canadian Centre for Justice Statistics, when the murderer and the victim are strangers, the murderer is extremely likely to be a male (96 percent of the cases) and is younger than 26 years old in half of the cases. Murderers who are younger than 18, as in about eight percent of cases, tend to beat or strangle their victims to death; this happened in 33 percent of murders committed by youths between 1991 and 1993. The young murderers stabbed or shot their victims 29 percent of the time during the same period (Wright and Federowycz, 1996: 71).

Between 1961 and 1990, female murderers accounted for only 12 percent of all murders in Canada and, in three-quarters of these cases, the victim was a family member. Female murderers are less likely to use a firearm than male offenders; they did so in 23 percent of cases (Silverman and Kennedy, 1993: 141).

During the 30-year period ending in 1990, about 71 percent of firearm-related homicide victims were male. In fact, when the victim of a homicide was a female, a firearm was somewhat less likely to be used than when the victim was male (Silverman and Kennedy, 1993).

When the murder victim is female, she is as much as nine times more likely to be killed by a spouse or by someone whom she has known intimately than by a stranger (Rodgers and Kong, 1996; Wilson et al., 1995; Wright and Fedorowycz, 1996: 68). In the 30 years between 1961 and 1990, 2,129 husbands killed their wives. They used a firearm 47 percent of the time, and even more frequently when they were older than 65 (Silverman and Kennedy, 1993: 69-76). Where the murderer was male, he then committed suicide 27 percent of the time, compared to three percent of female offenders (Ibidem). Among the 782 wives who killed their husbands during the same period, 35 percent used a firearm.

Silverman and Kennedy (Ibid.) point out that 41.7 percent of victims of firearm-related homicides between 1961 and 1990 were between 18 and 34 years old, and 40.7 percent were between 35 and 54 years of age. Victims who were younger than 18 accounted for 9.4 percent of all victims, and those over the age of 55 represented eight percent.

When the murder victim is an infant or a child, the murderer is often a parent. In 620 cases of a parent killing a child in Canada between 1961 and 1990, 323 involved the fathers and 289 involved the mothers. Mothers killed their children with a firearm in nine percent of cases; fathers 25 percent of the time. Parents were more likely to kill their infants and young children than their older children. In 43 percent of cases involving the mother, the child was less than two years old. When examining the method used, as the age of the victim increased, firearms were used more often (Silverman and Kennedy, 1993).

These statistics show how complex the question of homicides can be. In the last ten years, about 2,100 people were murdered by a firearm. When looking at how many of these could have been prevented, there are unresolved questions. For example, there were fewer firearm homicides in Canada in the last decade, while there were no significant changes in the number of firearms that were available in the country (Department of Justice Canada, 1996). And the reduced firearm homicide rate cannot explained by measures that made handguns and other types of firearms less accessible; the proportion of homicides involving handguns has been increasing.

Some answers may be found in the changing patterns of different types of homicides. Changes in patterns of armed robberies, for example, may affect the number of secondary homicides–those committed during another criminal offense.

5.3 Type of Firearms Used in Homicide

Homicide data from 1974 to 1996 indicate that 55 percent of all firearm homicides were committed with rifles and shotguns. However, since 1991, offenders have increasingly tended to murder their victims with a handgun, and have done so less often with a rifle or shotgun (Hung, 1997). Axon and Moyer (1994) examined homicides that occurred in Toronto from 1991 to 1993. In cases where they knew what type of firearm was used, 72 percent of the time it was a handgun; rifles or shotguns were used 20 percent of the time; and a sawed-off long gun was used in seven percent of cases. The percentage of cases involving handguns in this study was higher than the national average.

In 1996, among all firearm homicides:

  • 50 percent were committed with handguns;
  • 39 percent involved a rifle or shotgun; and
  • 11 percent involved a fully automatic firearm, a sawed-off rifle or shotgun, or an unknown type of firearm (Hung, 1997).

5.4 Firearms and Domestic Violence

In recent years, increased attention has been given to family homicide, and to spousal homicides in particular. Between 1975 and 1990, firearms were involved in approximately one-third of all domestic homicides (Dansys Consultant, 1992).

When there is a history of fights in the home, or when one or more members of the household abuses illegal substances, there is a greater risk of homicide. Research evidence shows that having a firearm in the home is also associated with a higher risk of homicide by a family member or intimate acquaintance (Boyd, 1995; Gabor, 1994; Kellermann et al., 1993). According to Reiss and Roth (1993: 262), the choice of a weapon in violent domestic disputes may well be "the nearest available object that can project force." In contrast to other types of homicide, the authors concluded, it would seem likely that in domestic disputes "the instrumentality rather than intent contributes most of the firearm’s lethal effect" (Ibidem).

Different strategies may be called for to prevent homicide in the home than to prevent homicide on the street (Tardiff et al., 1995). The murder of one spouse by another is usually preceded by other violent incidents that are often known to the police. Domestic homicides should be preventable, at least in some cases, by reducing the likelihood that a firearm is present during such conflicts.

Crawford and her colleagues studied intimate femicides in Ontario between 1991 and 1994. They found evidence that "intimate femicides were not the isolated and unpredictable acts of passion they are often believed to be" (1997: 50). In one-half of the cases, the offender had previously attacked or threatened the victim, and in at least one-third of the cases, the couple had had some contact with the police before the killing.

Prohibition orders and, to a lesser extent, measures to ensure the safe storage of the firearms that are kept at home, are other means that have been proposed to prevent incidents of domestic violence (Department of Justice Canada, 1995a). The effectiveness of such measures in preventing spousal homicides has not been empirically assessed. It is also unlikely that the measures in question can affect situations where the firearm used was obtained or possessed illegally by the offender. According to Dansys Consultants (1992: 26), as much as one-fifth of the firearms used in spousal homicide cases may fall within this category.

5.5 International Comparisons

International studies tend to show a positive correlation between levels of firearm ownership and homicide rates, even if the relationship is not exact (Gabor, 1994:35; 1995: 199), and indicate a strong statistical association between gun ownership levels and gun-related homicides (Killias, 1993b). Yet, the observed presence in some cases of a positive correlation between firearm ownership and non-firearm homicides suggests that other factors are at play in producing the observed correlations.

While there may be many reasons behind the different homicide rates in Canada and the United States, a comparison strongly suggests that the difference in the amount of available firearms in the two countries is an important factor. A recent analysis conducted for the Department of Justice Canada (Hung, 1996) revealed the following:

  • On average, between 1985 and 1995, the per capita homicide rate in the United States was 3.8 times higher than in Canada.
  • For the same period, twice as many homicides involved a firearm in the United States than in Canada.
  • In 1995, the U.S. per capita firearm homicide rate was 9.7 times higher than in Canada.

The killing of police officers in the line of duty is another area that illustrates the difference between firearm violence in the two countries. According to Gabor (1997: 12), when the relative number of sworn officers in the two countries is taken into account, a U.S. police officer is seven times more likely to be killed than a Canadian officer. In the United States, out of the 74 police murders which occurred in 1995, 83.7 percent involved the use of a firearm; it was a handgun in 58.1 percent of cases (Federal Bureau of Investigation, 1997).

5.6 Youth and Firearm Crime

5.6.1 Youth as Perpetrators of Firearm Homicides

From 1985 to 1992, the rates for homicides committed by children and youth remained much lower in Canada than in the United States; the rates have not increased since then (Silverman and Kennedy, 1993: 164). According to Moyer (1996:95), for homicide and attempted murder, the number of suspects between the ages of 12 and 17 fluctuated in the past 17 years, with no consistent trend. Between 1961 and 1990, children under the age of 18 committed 794 homicides (Silverman and Kennedy, 1993: 162). Firearms were used in 45 percent of cases involving children under the age of 15, and in 31 percent of cases involving an offender between 15 and 17 years of age (Ibidem).

5.6.2 Firearm-related Youth Crime

The Federal/Provincial/Territorial Task Force on Youth Justice concluded a review of violent youth crime by stating that there were clear differences between public perceptions of youth crime in Canada and reality (1996: 18) (See also: Schissel, 1997; Roberts, 1994: 46). Their report argued that the public is "undoubtedly influenced by the American media and popular culture, and is probably unaware of the very large differences between Canada and the United States in the amount and seriousness of violent youth crime" (Idem: 17). Similar misconceptions, based on the U.S. experience with youths using firearms and becoming violent, may also affect how serious the Canadian public perceives the problem of youth violence involving firearms to be. To date, there is no consistent evidence that more youths are using firearms in violent incidents today than in the last 20 years.

The public is concerned about the situation in the United States, where violent youth crime has reached alarming proportions. Although such incidents seem to have decreased in the last few years, the decrease is in relation to record-high rates in the previous decade. For example, the number of juveniles under 18 years of age who committed homicides doubled between 1985 and 1992. Beginning in 1985, the number of firearm homicides grew steadily, with no corresponding upward trend in homicides not involving a firearm (Bilchik: 1996; Blumstein: 1995; 1996; Blumstein and Cork, 1996; Cornell, 1993; Donzinger, 1996; Kellermann: 1995; Powell et al., 1996; Zimring, 1996).

The rates of youth crime in Canada remained much lower than in the United States during the same period. The rate of juveniles charged with using a firearm, imitation firearm or air gun compared to other types of robberies remained low (Moyer, 1996). There is little research on youths who participate in violent gangs (e.g., Mathews, 1993) and even less on how these groups use firearms. Some anecdotal evidence suggests that members of youth gangs in Canada are less likely than their U.S. counterparts to carry a firearm as opposed to another kind of weapon.

Moyer noted that the increase in violent youth crime in the United States occurred despite a toughening of juvenile justice legislation in some states. It was therefore most likely associated with social factors, and not legislation (Moyer, 1996). In fact, it would appear that one of the most significant factors in the levels of youth violence between the two countries may indeed be the ease, or lack of it, with which youths from the two countries can get firearms and, in particular, can obtain handguns.

5.6.3 The Diffusion Hypothesis

Several U.S. authors explain the unprecedented increase in youth violence involving firearms as the diffusion or contagion hypothesis (Bilchik, 1996; Blumstein, 1995; 1996; Blumstein and Cork, 1996; Travis, 1997). According to this hypothesis (Blumstein, 1996), juveniles who became increasingly involved in the drug trade acquired firearms to protect themselves; a relatively easy task, given the availability of firearms in the United States. Then, many youths not involved in the drug business may have felt it necessary to have a firearm to protect themselves from armed drug dealers. The increased presence of guns in the community has meant that disputes once settled by fights escalated to more lethal incidents involving shootings (Zimring, 1996). The end result, Blumstein observed, is that "gun possession escalated into an arms race that diffused the weapons broadly throughout the community" (1996: 2).

With this hypothesis many researchers conclude that, for U.S. youth, acquiring a firearm and committing violent crimes with firearms is no longer as closely related to drug trafficking as it once was (Kennedy et al., 1996: 153). Fear, self-protection and self-defence emerge as overwhelming reasons why a large proportion of U.S. youth, particularly in cities, have taken to carrying concealed weapons on a regular basis (Bilchik, 1996; Hemenway et al., 1996; Sheley and Brewer, 1995; Sheley and Wright, 1995). This is particularly true of youths involved in crime or in gangs (Ash et al., 1996; Bjerregaard and Lizotte, 1995; Callahan et al., 1993; Decker et al., 1996; Hutson et al., 1994; Kennedy et al., 1996; Koper and Reuter, 1996; Sheley and Wright, 1993; 1995).

At the centre of the diffusion hypothesis is the fact that the youths have easy access to firearms, particularly to handguns (Blumstein and Cork, 1996; Zimring, 1996). It is illegal in both Canada and the United States for a youth to possess or carry the type of firearm that Canada defines as restricted or prohibited. In contrast to the United States, however, where possession and carrying of such firearms is widespread among youths and particularly among youths involved in crime (Callahan et al., 1993; Decker et al., 1996; Kellermann, 1995; Sheley and Wright, 1993;1995), there is no evidence that this is the case in Canada. In this country, most cases of youths charged with possession of offensive weapons did not have handguns or prohibited weapons, such as switch blades, martial arts items or automatic firearms, but instead were carrying such things as bats, knives and sticks. In fact, incidents involving restricted weapons have been fairly rare over the last decade (Moyer, 1996: 100).

5.6.4 Violence in Schools

There are significant differences between Canada and the United States with respect to violence in schools and the prevalence of firearms in schools. A two-year U.S. study of 25 states, found that 105 people were killed in school incidents, and the offender used a firearm in 77.1 percent of the cases. About 95.6 percent of victims were male, and 72 percent were students. Victims were more likely to belong to a minority racial or ethnic group in a secondary school within an urban school district (Kachur et al., 1996; see also: Sheley et al., 1995; 1994a).

To date, the level of school violence observed in the United States is unmatched in Canada. Nevertheless, Canadian teachers, school board representatives and law enforcement officials have indicated in local surveys that they are concerned about an increase in the amount of violence in schools and in the number of students who carry weapons to school (e.g., Walker, 1994).

A 1995 survey of Canadian school board and police representatives indicated that 80 percent of respondents believed that violence was more common and intense than ten years ago (Gabor, 1995). A national mail-out survey of police officials and educators indicated that it is rare to find and seize firearms in junior and senior high schools and that seizures tend to take place in urban centres of 50,000 or more people. Most weapons seized were knives, shop-crafted or homemade weapons and clubs, bats and sticks. The use of weapons in violent confrontations between youths in schools was not believed to be common (Walker, 1994: 8).

While police seizures of weapons may be relatively rare in Canada, another study indicates that weapons may be more present in urban high schools. A 1995 survey of 962 Calgary secondary school students found that 28 percent of respondents admitted to carrying a weapon at school or having a weapon in their lockers during the past year. The weapon in question was most often a knife, at 15.9 percent, a homemade weapon, at 11.6 percent, or a club or bat, at 9.1 percent. With respect to firearms, the students had handguns 2.6 percent of the time; followed by pellet guns 5.1 percent of the time, and replica firearms 6.5 percent of the time (Smith et al.: 1995; 1995a). Four out of five students who had brought a handgun to school–mostly male students–reported doing so once or a few times (Smith et al., 1995a: 60).

Date modified: