You are hereSkip Navigation Links > Home > Professionals > Ontario's Youth Action Plan > Roots of Youth Violence > Volume 4

Download print version (PDF).

Review of the Roots of Youth Violence: Research Papers

Volume 4

A Province at the Crossroads: Statistics on Youth Violence in Ontario

A Report Prepared for the Review of the Roots of Youth Violence



Dr. Scot Wortley1
Associate Professor
Centre of Criminology, University of Toronto

June 2008

1. The author can be reached at the Centre of Criminology, University of Toronto, 14 Queens Park Crescent West, Toronto, Ontario, M5S 3K9; phone: 416-978-7124 (ext. 228); email: scot.wortley@utoronto.ca.

Introduction


On December 26, 2005, Jane Creba, a 15-year-old student out exploring Boxing Day sales with her mother, was shot to death in the middle of Toronto’s busiest shopping district. Six other innocent bystanders were wounded. Creba and the other victims had apparently been caught in the crossfire of a shootout between two rival youth gangs. On May 23, 2007, Jordan Manners, another 15-year-old Toronto student, was shot and killed within a local high school during school hours. The school was locked down for several hours while the police searched for the killers. Both of these tragedies received national media attention and became topics of conversation among politicians, law enforcement officials, educators, community workers and the general public. These two cases, however, should not be viewed as isolated incidents. The following is a list of other recent, but less famous, homicides involving young Ontario victims:

The names of so many young homicide victims can bring cold crime statistics to life. At an emotional level, they capture the reality of violence in our society. Unfortunately, the list of names provided above is far from exhaustive. Over the past decade, many other young men and women from Ontario have been murdered and countless others have been the victims of nonlethal forms of violence, including physical assault, robbery, and sexual assault. It is also unfortunate that, unlike Jane Creba and Jordan Manners, most young victims, including those listed above, will not become household names. Indeed, it would be safe to say that the general public has already forgotten many of the cases listed above. This, however, does not make these incidents any less tragic. Indeed, all violence results in pain and heartbreak for the families, friends and communities involved.

A focus on individual cases, like those listed above, can also create fear of crime and contribute to the belief that violent crime is getting worse. Criminologists, however, know that this is not necessarily true. Violent crime has always existed in Canada and each decade has seen its share of sensational incidents. The following Toronto examples, from the 1960s and 1970s, serve to illustrate this point:

At the time, all of these cases produced sensational headlines and calls for a tougher criminal justice system. They also produced concerns that violent crime was increasing rapidly in Canada and that Toronto was now subject to “American-style” violence. Nothing much has changed over the past 30 years. Media coverage of individual cases still has a powerful impact on public perceptions of crime and violence. Criminologists, on the other hand, tend to rely on both official and unofficial crime statistics when forming their opinions. Aggregate statistics are far superior to individual case descriptions when it comes to documenting overall crime patterns and trends. In this report, therefore, both official and unofficial crime statistics are used to address the following six research questions:

  1. How prevalent is violent behaviour in Ontario?
  2. What proportion of violent crime in Ontario involves young people?
  3. Is Ontario more or less violent than other Canadian provinces?
  4. Are Ontario cities more or less violent than cities located in other Canadian provinces?
  5. Does Ontario have more or less violence than the United States and other foreign countries?
  6. Has violent offending in Ontario increased or decreased over the past 40 years?

The report begins with a discussion of how social scientists measure violent crime. We then review official police statistics on violent crime before turning to estimates derived from both victimization and self-report surveys. The report concludes by highlighting several disturbing trends with respect to youth violence in Ontario.

Measuring Violent Crime with Official Data

Measuring the true level of crime and violence within a given society is a very difficult task. Typically, criminologists, law enforcement officials, and policy-makers rely on two major categories of crime data: 1) official police statistics; and 2) unofficial survey data (self-reports about personal victimization experiences or criminal behaviours). As discussed below, both forms of data collection have their strengths and weaknesses. Thus, crime experts generally concede that we must closely examine both official and unofficial sources of crime data in order to develop a fuller understanding of the nature and extent of criminal behaviour. We begin our review of crime in Ontario with an examination of police crime data. We then turn our attention to survey data.

The Uniform Crime Reporting Survey (UCR)

Through the use of the Uniform Crime Reporting Survey (UCR), the Canadian Centre for Justice Statistics (CCJS) has collected information on crimes recorded by the police in Canada since 1962. In 2006, survey respondents included over 400 different police services from across the country — the vast majority of police services at the municipal, provincial and national levels. Indeed, according to recent estimates, in 2006 police service coverage was at 99.9 per cent of the national caseload. After receiving crime data from individual police services, the information is ultimately aggregated by the CCJS and released to the public on an annual basis.

How Are UCR Incidents Counted?

The methodology used to collect UCR crime data is rather complex. Every month, individual police services compile the total number of crimes recorded in their jurisdictions (by crime type) and report these incidents to the Canadian Centre for Justice Statistics. Crime counts are based on all criminal incidents reported to the police by members of the public and all crimes that are discovered by the police through routine patrol or proactive investigation. In other words, UCR crime data reflect only those crimes that are “known to the police.” Previous research suggests that many crimes are, in fact, not reported to the police by the public (see discussion in the next section) and thus go undetected. Thus, there is a general consensus among criminologists that official UCR data significantly underestimate the true level of crime in society. However, experts also agree that some UCR data are more accurate than others. For example, research suggests that while most physical assaults are never reported to the police, law enforcement officials do detect the vast majority of homicides. Thus, while UCR homicide data are considered very accurate, it is conceded that police assault data only document a fraction of the physical assaults that take place in Canada within any given year. As a general rule, the more serious the crime (in terms of physical injury or financial loss), the more likely it is to be recorded in UCR statistics.1

Criminal incidents are recorded individually and divided into four major categories: 1) violent crime; 2) property crime; 3) criminal traffic violations; and 4) drug-related crimes. The number of criminal incidents recorded by the UCR represents “actual” crime. In other words, the survey only records crimes that have been substantiated by police. Unfounded reports are not included in “actual” crime counts.

An important methodological note, however, concerns incidents in which multiple criminal offences have taken place. When there are multiple offences during a single incident, the UCR only records the most serious offence. For example, if a single incident involved both a murder and a robbery, only the murder would appear in UCR statistics. According to UCR protocol, violent offences are always considered more serious than property offences.2

Violent offences are counted with respect to the total number of crime victims. Property crimes, on the other hand, are counted only once per occurrence. For example, if four individuals were physically assaulted at one time, the event would be counted as four different assaults (i.e., four violent incidents). By contrast, a break and enter case would be recorded as a single criminal incident — even if many victims had lost their property. The exception to this general rule is robbery. Although the UCR records robbery as a violent crime, robberies are always counted as a single incident — even in cases where there were multiple victims.

After police investigation, some reported criminal incidents are determined to be unfounded or false. These cases are always eliminated from UCR crime counts. It should be noted, however, that the UCR records all “confirmed” criminal incidents — even if an offender is never identified or property is never recovered.

The Revised UCR

In 1984, the UCR was revised so that it could collect more detailed information about criminal incidents. For example, in contrast to the original UCR, the UCR2 tries to collect detailed information on victim and offender characteristics (age, gender, alcohol and drug use, gang involvement, etc.), the nature of the victim-offender relationship (family member, friend, stranger, etc.), and the nature of the criminal incident (time, location, extent of injury, amount of property loss, etc.) By 2006, almost all of the police services in all provinces and territories were using the UCR2 (except British Columbia). Data from the UCR2 is aggregated at year-end to maintain compatibility with the original UCR (see Silver 2007).

Expressing Crime Data: Crime Counts and Crime Rates

The UCR uses a variety of methods to express or communicate Canadian crime statistics. First of all, the total number of crimes reported and the total number of arrests made by the police are provided as raw numbers. For example, according to UCR data, Canada experienced 605 homicides in 2006. Secondly, the percentage change in the amount of crime between years is computed. For example, the total crime rate in Canada decreased by three per cent between 2005 and 2006. The percentage change calculation provides some indication of whether crime is increasing or decreasing and can thus be used to identify short-term crime trends.

The third and by far the most important way of summarizing crime data is through the calculation of crime rates (per 100,000 population). In order to calculate a crime rate, one must first divide the total number of crimes by the population estimate for the region (typically derived from census data). The result is then multiplied by one hundred thousand in order to produce a crime rate. Let us consider the calculation of Canada’s homicide rate for 2006. The formula would look something like this:

Total # of homicides in 2006/total Canadian population × 100,000 = homicide rate

Canada experienced 605 homicides in 2006. Census estimates put Canada’s total population at 31,372,587 individuals:

605 homicides/31,372,587 × 100,000 = 1.85

What does the crime rate mean? How can it be interpreted? As calculated above, in 2006, Canada had a homicide rate of 1.85 per 100,000 people. This means that, in Canada, between January 1 and December 31, 2006, approximately two out of every 100,000 people were murdered.

One of the values of crime rates (as opposed to crime counts) is that they allow us to more accurately compare the level of crime in society over time and across jurisdictions. For example, in 1961, Ontario experienced 89 homicides. By 2006 the number of homicides in the province had climbed to 196. These figures represent a 120 per cent increase in the total number of homicides in Ontario over a 35-year period. Unfortunately, a casual observer might look at these figures and wrongly conclude that Ontario residents were twice as likely to be murdered in 2006 than 1961. The problem is that these raw homicide numbers do not take population growth into consideration. Indeed, between 1961 and 2006, Ontario’s population grew from approximately 6.2 million to approximately 12.7 million (an increase of 105 per cent). Thus, a comparison of homicide rates is much more appropriate with respect to documenting whether Ontario is becoming more dangerous. In fact, between 1961 and 2006, Ontario’s homicide rate increased only slightly, from 1.43 per 100,000 to 1.54 per 100,000 (an increase of only eight per cent).

Homicide rates also permit more accurate regional comparisons. For example, in 2006, Ontario experienced 196 homicides while Saskatchewan recorded only 40 homicides. This might cause some to wrongly conclude that Ontario is much more violent or dangerous than Saskatchewan is. After all, Ontario recorded almost five times more homicides than Saskatchewan did. Once again, however, these raw homicide numbers have not controlled for population size. Indeed, while Ontario has 12.7 million residents, the population of Saskatchewan has less than a million (only 985,000 residents). In fact, according to an analysis of homicide rates, Saskatchewan is actually a much more dangerous province (homicide rate = 4.06 per 100,000) than Ontario is (homicide rate = 1.54 per 100,000). It is for these reasons that most of the following discussion of crime statistics will focus on crime rates rather than raw crime numbers.

The Extent of “Youth Crime”

The level of “youth violence” in a particular society will largely depend on one’s definition of “youth.” In Canada, we typically distinguish between crimes committed by young offenders (17 years of age and under) and crimes committed by adults (18 years of age or older). However, while the behaviour of legally defined “young offenders” is a definite focus of the Review of the Roots of Youth Violence, we are also quite interested in violence that involves young adults (between 18 and 29 years of age).

Unfortunately, the extent of crime and violence committed by people from different age groups is very difficult to determine in Canada. First of all, unless a person is actually charged with a crime, the police will not know — and thus will not be able to record — the age of the offender. This means that the age of the offender is actually “unknown” in a very high proportion of criminal incidents. For example, in 1999, the age of the perpetrator(s) was recorded for only 33 per cent of the 1.17 million criminal incidents documented by Canada’s UCR2 survey (Carrington 2001).3

Furthermore, while arrest statistics for young offenders (12 to 17 years of age) are regularly provided by Statistics Canada, detailed age breakdowns for adult offenders are rarely released to the public. This makes it particularly difficult to calculate age-specific crime rates or identify the proportion of arrests in Canada that involve young adults (18 to 29 years).4

Nonetheless, there is a huge body of international research that documents that there is a strong, inverse relationship between age and crime. Regardless of gender, ethnicity and socio-economic status, cross-national research suggests that younger people are much more likely to engage in crime and violence than older people are. This pattern has remained remarkably stable over the past century and is evident in all developed and developing nations (see Siegel and McCormick 2006; Carrington 2001). In general, rates of criminal offending and violent behaviour are highest among those in their mid-teens to early twenties. By contrast, criminal behaviour, especially violent criminal behaviour, declines significantly in the late twenties and thirties and is increasingly uncommon among people over 40 years of age. This phenomenon has come to be known as the “age-crime” curve.

After conducting a complex series of calculations and estimates, Carrington (2001) produced age-crime estimates for Canada using 1999 UCR data. The results powerfully illustrate that criminal activity is highly concentrated among adolescents and young adults (see Figure A). For example, 15–24-year-olds represented only 14 per cent of the Canadian population in 1999, but accounted for over a third (36 per cent) of all criminal incidents recorded by the UCR. By contrast, Canadians 50 years of age or older represent 28 per cent of the population but accounted for only seven per cent of recorded criminal activity.

Percent of Canadian Population and Proportion of Recorded UCR Crimes by Age Group, 1999 Estimates (from Carrington 2001)
 Figure A: Percent of Canadian Population and Proportion of Recorded UCR Crimes by Age Group, 1999 Estimates (from Carrington 2001)

Carrington (2001) also calculated age-specific crime rates using 1999 UCR2 data. The results of this analysis further illustrate the inverse relationship between age and criminal activity (see Figure B). The highest crime rate (25,708 crimes per 100,000) was recorded for 15–19-year-olds, followed closely by 20–24-year olds (20,022 per 100,00) and 25–29-year-olds (15,363 per 100,000). The crime rate for 15–24-year-olds is three times higher than the national average (8,750 per 100,000). By contrast, crime rates decline dramatically after 40 years of age.

Graph: Estimated Age-Specific Crime Rates (per 100,000), 1999 Canadian UCR Data (from Carrington 2001)
Figure B: Estimated Age-Specific Crime Rates (per 100,000), 1999 Canadian UCR Data (from Carrington 2001)

Unfortunately, we could not locate age-specific UCR crime statistics for Ontario. The Canadian Centre of Justice Statistics has yet to release these data. However, the Review of the Roots of Youth Violence was able to obtain age-specific criminal charge data from the Ontario Attorney General. The data indicate that, in 2007, the police in Ontario laid 569,072 different criminal charges. One out of five criminal charges (22.1 per cent) involved a violent offence (homicide, attempted murder, aggravated assault, assault, sexual assault and robbery). Figure C breaks down these criminal charges by age group. The data indicate that young people in Ontario are grossly over-represented among those who were charged with a criminal offence in 2007. For example, although 12–17-year-olds are only eight per cent of Ontario’s population, they represent 14 per cent of all those charged with a criminal offence and 15 per cent of those charged with a violent crime. Similarly, 18–24-year-olds are only nine per cent of Ontario’s population, but represent 28 per cent of those charged with a criminal offence and 24 per cent of those charged with a violent crime. In sum, although they represent only 23 per cent of the Ontario population, 56 per cent of those charged with a crime in 2007, and 52 per cent charged with a violent crime, were between 12 and 29 years of age. By contrast, those 60 years of age and over represent 17 per cent of Ontario’s population, but only two per cent of those charged with a criminal offence. These data serve to further illustrate the strong negative relationship between age and criminal behaviour.

Graph: Percent of Population and Percent of all Criminal Charges by Age Group, Ontario 2007
Figure C: Percent of Population and Percent of all Criminal Charges by Age Group, Ontario 2007

Figure D presents 2007 violence-related criminal charge rates by age group. The basic pattern that emerges is remarkably consistent with the classic age-crime curve discussed above (see Figure B). The results indicate, for example, that 18–24-year-olds have by far the highest rate of violent crime in Ontario (2,824 per 100,000), followed by 25–29-year-olds (2,169 per 100,000) and those between 12 and 17 years of age (1,904 per 100,000). The rate of violent crime, by contrast, declines dramatically after 30 years of age. Indeed, the violent charge rate for 18–24-year-olds is 21 times greater than the violent charge rate for Ontario residents 60 years of age or older.

Graph: Violence-Related Criminal Charge Rate (per 100,000) by Age Group, Ontario 2007
Figure D: Violence-Related Criminal Charge Rate (per 100,000) by Age Group, Ontario 2007
The following crime statistics further serve to illustrate the negative relationship between age and violence in Canada:5

In conclusion, the data described above strongly suggest that, when we are talking about crime data, we are most often talking about the behaviour of young people (typically between 12 and 29 years of age). Thus, we feel that the aggregate crime data presented in the next section are reflective of youth crime trends in general. Indeed, the negative relationship between age and crime is so strong that several scholars have forecast significant declines in North American crime rates as the population ages (see Field 1999; Savolainen 2000; Carrington 2001).

Violent Crime: Provincial Comparisons

On an annual basis, the Canadian Centre for Justice Statistics (CCJS) releases crime statistics for Canada. At the time that this report was being prepared (April 2008), statistics for 2006 were available for analysis. Unfortunately, the figures for 2007 had yet to be released. Figure 1 presents the rate of violent crime for Canada’s ten provinces. The violent crime rate is a composite measure of the following crimes: homicide, manslaughter, attempted homicide, physical assault, aggravated assault, assault with a weapon, sexual assault and robbery. In general, the data suggest that the Western provinces have a much higher rate of violent crime than the eastern provinces do (with the possible exception of Nova Scotia). Over all, at 756 violent crimes per 100,000, Ontario and Quebec have the second-lowest violent crime rate in Canada. Only Prince Edward Island reported less violent crime (714 per 100,000). By contrast, the rate of violent crime in both Saskatchewan (2,039 per 100,000) and Manitoba (1,598 per 100,000) is more than double the rate for Ontario.

Graph: Figure 1: 2006 Violent Crime Rates (per 100,000) by Province
Figure 1: 2006 Violent Crime Rates (per 100,000) by Province

Figure 2 presents the homicide figures by province. Consistent with the figures for overall violent crime, the Western provinces (Saskatchewan, Manitoba, Alberta and British Columbia) have significantly higher homicide rates than the Eastern provinces have. Ontario’s homicide rate (1.5 per 100,000) is half the rate of Saskatchewan’s (4.1 per 100,000) and Manitoba’s (3.3 per 100,000). It is also significantly lower than the homicide rate for both Alberta (2.8 per 100,000) and British Columbia (2.5 per 100,000). However, Ontario’s homicide rate is significantly higher than the homicide rates for Quebec, New Brunswick and Prince Edward Island.

Graph: Figure 2: 2006 Homicide Rates by Province
Figure 2: 2006 Homicide Rates by Province

Figures 3 through 5 present the provincial rates for assault, robbery and sexual assault respectively. The results are remarkably consistent. Regardless of the type of crime, violence is far more prevalent in Western Canada than it is in Eastern Canada. Ontario, on the other hand, is consistently at the low end of the crime continuum. Indeed, in 2006, Ontario recorded the second-lowest rate of physical assault (563 per 100,000) in Canada and the second-lowest rate of sexual assault (56 per 100,000). By contrast, Saskatchewan’s physical assault rate (1,671 per 100,000) is almost three times greater than Ontario’s assault rate. Similarly, Saskatchewan’s rate of sexual assault (125 per 100,000) is more than twice Ontario’s rate. However, the data suggest that Ontario has a relatively high rate of robbery (87 per 100,000). Nonetheless, Ontario’s robbery rate is still lower than the rate found in five other provinces, including Manitoba (182 per 100,000), Saskatchewan (150 per 100,000), British Columbia (110 per 100,000), Alberta (93 per 100,000) and Quebec (91 per 100,000). In sum, compared with other provinces in Canada, Ontario is a relatively safe place to live.

Graph: Figure 3: 2006 Assault Rates by Province
Figure 3: 2006 Assault Rates by Province

Graph: Figure 4: 2006 Robbery Rates by Province
Figure 4: 2006 Robbery Rates by Province

Graph: Figure 5: 2006 Sexual Assault Rates by Province
Figure 5: 2006 Sexual Assault Rates by Province

Violent Crime: A Comparison of Canadian Cities

Over the past several years, high-profile murders in Toronto (Canada’s largest urban area) have drawn national media attention. As a result, many Canadians believe that Toronto and other Ontario cities are particularly prone to violence (see MacQueen 2008). For example, a general population survey, conducted in 2007, revealed that 50 per cent of Toronto residents actually believe Toronto has more crime than other major cities in Canada (Wortley 2007). How accurate is this perception? An examination of 2006 UCR statistics reveals that it is not accurate at all. Figure 6, for example, presents the violent crime rates for 20 major urban areas in Canada, including ten urban areas located in Ontario (Toronto, London, Hamilton, Windsor, St. Catharines, Ottawa, Thunder Bay, Sudbury, Kingston and Kitchener-Waterloo). According to our analysis, only two of these Ontario cities — Thunder Bay (1,308 per 100,000) and Sudbury (908 per 100,000) — make it into the top ten most violent cities in Canada. Interestingly, these are also the only two cities in Ontario with a violent crime rate above the national average (951 per 100,000). Toronto, on the other hand, a city that is often stereotyped as violent, ranks 14th out of the 20 urban areas in our sample. Furthermore, Toronto’s violent crime rate (712 per 100,000) is less than half the rate of Saskatoon’s (1,606 per 100,000) and Regina’s (1,546 per 100,000). Toronto’s violent crime rate also falls far below the national average. Nonetheless, there are several other Ontario cities (Ottawa, St. Catharines, Kitchener-Waterloo, Windsor and Kingston) that have even less violence than Toronto has. Consistent with our analysis of provincial data, Western cities tend to have much higher levels of violent crime than Eastern cities have. Halifax, Nova Scotia, is a notable exception to this general trend.

Graph: Figure 6: 2006 Violent Crime Rates by Canadian City
Figure 6: 2006 Violent Crime Rates by Canadian City

Graph: Figure 7: 2006 Homicide Rates by Canadian City
Figure 7: 2006 Homicide Rates by Canadian City

Figure 7 presents homicide statistics and Figure 8 presents robbery statistics for the same 20 Canadian cities. Some argue that homicide and robbery rates provide a more accurate estimate of violent crime because these crimes are more likely to be discovered by the police than physical assaults or sexual assaults are. Furthermore, compared with assault figures, homicide statistics are less vulnerable to regional variations in police discretion. For example, although local police may decide not to record a minor assault (or decide not to lay a charge), they most certainly will record all homicides. The results with respect to homicide and robbery, however, are remarkably similar to the results for violent crime in general. Over all, Western cities have much higher homicide rates than Eastern cities have. For example, of the major cities in our sample, Regina has the highest homicide rate (4.5 per 100,000), followed by Edmonton (3.7 per 100,000), Saskatoon (3.3 per 100,000) and Winnipeg (3.0 per 100,000). By contrast, in 2006, the highest homicide rate in Ontario (1.8 per 100,000) was recorded by both Toronto and Ottawa. Nonetheless, compared with the level of homicide in Kitchener-Waterloo (0.4 per 100,000), St. Catharines (0.9 per 100,000) and Hamilton (1.0 per 100,000), Toronto’s homicide rate is relatively high — at least by Ontario standards.

Graph: Figure 8: 2006 Robbery Rates by Canadian City
Figure 8: 2006 Robbery Rates by Canadian City

With respect to robbery rates, only three Ontario cities — Toronto, Thunder Bay and Hamilton — make it into the top ten (see Figure 8). However, the robbery rates for these three Ontario cities (all between 108 and 116 per 100,000) are still far below the robbery rates recorded by Western cities like Winnipeg (272 per 100,000), Saskatoon (268 per 100,000) and Regina (264 per 100,000).

In sum, according to official police statistics, there is absolutely no truth to the perception that major Ontario cities — including Toronto — are more violent than other urban centres within Canada are. In fact, the violent crime rates for most Ontario cities fall far below the national average. On the other hand, most major Western cities (including Winnipeg, Regina, Saskatoon, Calgary, Edmonton, Vancouver and Victoria) have rates of violent crime well above the national average. This conclusion is quite consistent with a recent, highly publicized analysis of UCR data conducted by Maclean’s magazine. Maclean’s researchers ranked the 100 largest urban areas in Canada (with populations greater than 50,000) on an aggregate measure of six crimes: murder, sexual assault, aggravated assault, robbery, break and entering and auto theft. Surprisingly (at least to some), Toronto ranked 26th on this list of the most dangerous urban areas in Canada. By contrast, nine of the top ten most dangerous cities are located in Western provinces — from Winnipeg to Victoria (see MacQueen 2008; macleans.ca/dangerouscities).

Violent Crime: International Comparisons

Comparing countries with respect to their level of violent crime is an extremely difficult task. To begin with, most countries have their own unique definitions of crime, making accurate cross-national comparisons virtually impossible. Furthermore, police agencies from different countries often have vastly different practices and standards when it comes to recording crime statistics. For example, while countries like Canada and the United States have rather sophisticated methods (Uniform Crime Reports) for collecting information on criminal incidents, other countries do not compile crime statistics with such rigor. This situation makes cross-national comparisons with respect to crime even more tenuous. As a result, most criminologists rely exclusively on homicide statistics when attempting to compare the level of violence in various countries. As discussed above, unlike other violent crimes, almost all homicides eventually come to the attention of the police. In other words, few homicides go unrecorded. Furthermore, as a reliability check, police data on homicide can be cross-referenced with “cause of death” statistics compiled by national health agencies. In the past, such reliability checks have demonstrated that national homicide statistics are relatively accurate (see Archer and Gartner 1984; Bailey and Peterson 1999). Finally, many scholars feel that homicide is a “tip of the iceberg” statistic. They maintain that if homicide rates in a particular jurisdiction are high, it is reasonable to assume that less serious, non-lethal forms of violence will also be high.

Figure 9 compares Ontario’s 2006 homicide rate with recently compiled homicide statistics from a variety of developed and developing nations. The results clearly suggest that, compared with many countries, Ontario is a relatively safe place to live. In general, homicide rates are much higher in developing nations like Jamaica (62.1 per 100,000), Brazil (53.3 per 100,000) and South Africa (40.5 per 100,000), than in developed countries like Canada (1.8 per 100,000). Homicide rates are also high in Eastern Europe, particularly in those countries that used to belong to the Soviet Union. For example, the homicide rate in Russia (19.9 per 100,000) is approximately 13 times higher than the homicide rate in Ontario (1.5 per 100,000). By contrast, the homicide rate in the United States (5.7 per 100,000) is only 4 times greater than the homicide rate in Ontario. Interestingly, Ontario has a lower murder rate than many other Western nations, including Germany (2.9 per 100,000), Switzerland (2.7 per 100,000), Sweden (2.6 per 100,000), New Zealand (2.4 per 100,000) and Finland (2.1 per 100,000). However, Ontario is not doing as well as some other nations are. For example, Ontario’s homicide rate is three times higher than the homicide rate in Japan, Hong Kong and Singapore.

Figure 10 compares Ontario’s 2006 homicide rate with the homicide rates from selected American states. In general, the results suggest that Ontario is much safer than most jurisdictions in the United States. For example, the homicide rate in Louisiana (12.4 per 100,000) is approximately eight times greater than the homicide rate in Ontario (1.5 per 100,000). Florida, the winter home of many “snowbirds” from Ontario, has a homicide rate (6.2 per 100,000) that is more than four times greater than Ontario’s rate. Similarly, the homicide rates for New Jersey (4.9 per 100,000), New York (4.8 per 100,000) and Ohio (4.7 per 100,000) are more than three times greater than the rate for Ontario. While several states — including Hawaii, Maine, Utah and Iowa — have a rate of homicide that is similar to Ontario’s, we could locate only two states (North Dakota and New Hampshire) with lower rates.

As discussed above, in 2006, Toronto and Ottawa recorded the highest homicide rates among Ontario cities (1.8 per 100,000). Figure 11 compares this rate with 2006 homicide rates for selected American cities. The data clearly illustrate that, in general, urban living in Ontario is far safer than urban living in the United States is. In 2006, for example, Detroit had the highest urban homicide rate (47.3 per 100,000) in America. It is startling to note that this rate is 26 times greater than the homicide rate of Ontario’s “most dangerous” cities. It is also important to note that, despite cutting its homicide rate in half over the past 15 years, New York City’s 2006 homicide rate (7.3 per 100,000) is still four times greater than Toronto’s homicide rate. Even placid Salt Lake City has a homicide rate that is double the rate for Toronto and Ottawa. In fact, we could not locate a single American urban area (with a population greater than 250,000) with a lower homicide rate than the most “violent” cities in Ontario.

Figure 12 compares the homicide rates for Toronto and Ottawa with the homicide rates for selected European cities. The results suggest that, even when compared with European urban centres, Toronto and Ottawa are relatively safe. Indeed, Toronto and Ottawa (Ontario’s murder capitals) have lower homicide rates than many European locations have, including London, England, Glasgow, Amsterdam, Belfast, Brussels, Warsaw, Helsinki, Dublin, Copenhagen and Budapest. On the other hand, the homicide rates for Toronto and Ottawa are quite similar to the homicide rates for Berlin, Oslo, Madrid and Paris. However, a number of European cities (including Rome, Athens, Vienna and Lisbon) appear to be significantly safer than Toronto and Ottawa are. Indeed, Toronto’s homicide rate (1.8 per 100,000) is three times higher than the rate for Lisbon (0.6 per 100,000). Thus, although Ontario cities are relatively “safe” by European standards, theoretically we could do better.

Graph: Figure 9: Homicide Rates Ontario and Selected Countries
Figure 9: Homicide Rates Ontario and Selected Countries

Graph: Figure 10: 2006 Murder Rates Ontario and Selected American States
Figure 10: 2006 Murder Rates Ontario and Selected American States

Graph: Figure 11: 2006 Homicide Rates by Selected Ontario and American Cities
Figure 11: 2006 Homicide Rates by Selected Ontario and American Cities

Graph: Figure 12: Homicide Rates by Selected Ontario and European Cities
Figure 12: Homicide Rates by Selected Ontario and European Cities

Is Violence in Ontario Increasing?

The majority of Ontario residents believe that violent crime is on the rise. For example, the results of a 2007 general population survey suggest that over 70 per cent of Toronto residents believe that crime has increased significantly over the past ten years (Wortley 2007). The results presented below suggest that this widespread perception is fundamentally incorrect (see Figures 13 and 14).

We scoured annual reports published by the Canadian Centre for Justice Statistics and were able to locate the UCR violent crime rate for Ontario for the years 1986 through 2006. As you might recall, the violent crime rate is a composite measure that includes homicide, attempted homicide, assault, sexual assault and robbery. The data are presented in Figure 13. The results suggest that the violent crime rate in Ontario increased by 37 per cent between 1986 (787 per 100,000) and 1991 (1,097 per 100,000). However, violent crime in Ontario actually declined by 31 per cent between 1991 (1,097 per 100,000) and 2006 (756 per 100,000). Many Ontario residents would be surprised to learn that, over all, violent crime has actually decreased in this province over the past 20 years. Indeed, in Ontario, the 2006 violent crime rate (756 per 100,000) was five per cent lower than the rate of violent crime recorded in 1986 (797 per 100,000).

Graph: Figure 13: Ontario's Violent Crime Rate (per 100,000) 1986 to 2006
Figure 13: Ontario's Violent Crime Rate (per 100,000) 1986 to 2006

What about serious violence? Has it increased? In order to answer this question, we were able to locate homicide statistics for Canada, Ontario and the United States from 1961 to 2006 (a period of 45 years). These data are presented in Figure 14. The data illustrate a few clear patterns. First of all, over the past half century, the United States has consistently recorded a much higher homicide rate than Canada has. Depending on the year, the American homicide rate has typically been between three and six times higher than the Canadian rate. The results also reveal that, over the past 45 years, Ontario’s homicide rate has always been below the Canadian average. We could not locate a single year in which Ontario’s homicide rate surpassed the overall rate for Canada.

Ontario recorded its lowest homicide rate (1.02 per 100,000) in 1966. However, the province’s homicide rate more than doubled between 1966 and the mid-1970s. In fact, Ontario recorded its highest recorded homicide rate in 1975 (2.48 per 100,000). The level of homicide dropped slightly in 1976 and remained relatively stable (at about 2.0 per 100,000) until 1991. Ontario recorded relatively high homicide rates in both 1991 (2.35 per 100,000) and 1992 (2.29 per 100,000) before dropping below 2.0 per 100,000 for the next 14 years. As discussed above, in 2006, Ontario’s homicide rate was only 1.54 per 100,000. This rate is about the same as the homicide rate recorded in 1961 (1.43 per 100,000) and is actually 40 per cent lower than the rate recorded in 1975 (2.48 per 100,000). Many people would be surprised to know that Ontario’s 1975 homicide rate was almost twice the rate recorded in 2006. Clearly, there is little evidence to suggest that Ontario residents are more at risk of experiencing a violent death than they were 30 years ago. However, this does not mean that the nature of homicide has not changed. Indeed, there are a number of disturbing trends with respect to contemporary violence that deserve our attention. These trends will be discussed below in a section entitled “Ontario at the Crossroads.”

Graph: Figure 14: Homicide Rates, 1961 to 2006 United States, Canada and Ontario
Figure 14: Homicide Rates, 1961 to 2006 United States, Canada and Ontario

The Limitations of Official (UCR) Data

The UCR survey is an important crime measurement strategy that is used extensively by academics, community members and law enforcement agencies. However, police data are not without their problems. To begin with, UCR statistics only represent those crimes that are “known to the police.” Most crimes, as discussed above, become known to the police through reports from civilian witnesses and/or victims. However, survey results (see discussion below) indicate that many people do not report to the police the crimes they have witnessed or experienced. Reasons for not reporting crimes to the police sometimes include a belief that the incident was too minor or that the police will not be able to catch the offender. Other people do not report criminal incidents because they are afraid of the offenders or have a profound distrust of the police. Whatever the reasons, studies consistently reveal that there is a less than 50 per cent chance of a violent crime being reported to the police. This is even lower for certain types of violent crime, including sexual assault. In other words, official police data may dramatically underestimate the true extent of violent behaviour in Ontario.

Police discretion with respect to arrest decisions and record-keeping can also dramatically impact UCR statistics. Research suggests that police priorities and how they deal with specific types of behaviour can have a profound impact on the extent of criminal activity that is recorded by official statistics. For example, studies suggest that when the police devote extra resources towards identifying and arresting prostitutes, drug users, and drug dealers, the official statistics for prostitution, drug possession, and drug trafficking also increase. In other words, a rise or decline in particular “police sensitive” crimes may reflect changes in policing activity more than actual changes in the public’s behaviour. Indeed, in 1962, Canada recorded only 20 cases of cannabis possession. By 1968, however, there were 2,300 cases, and by 1972, there were over 12,000 cases (a 600 per cent increase over a ten year period). Although marijuana use may have increased somewhat over this time period, most police scholars attribute this dramatic increase in marijuana-related cases to renewed police efforts to fight the war on drugs (see Siegel and McCormick 2006).

A similar situation arose in Ontario during the 1990s. After the passage of the Safe Schools Act and the adoption of strict “zero tolerance” policies by various school boards, there was a marked increase in Ontario’s violent crime rate for young offenders. Most of this increase involved minor (Level I) physical assaults. Critics have argued that this increase in the province’s official violent crime rate had more to do with the increased use of police in schools than with real changes in youth violence. In other words, during the 1970s and 1980s, fights between students were often dealt with informally (by school officials, counsellors and parents). However, after the adoption of the zero-tolerance approach, the police were more likely to be called to schools to deal with minor violence. The increased use of police, in turn, led to increased arrests and an increase in officially recorded youth crime (see Doob and Cesaroni 2004).

Other studies have suggested that improvements in police record-keeping can also increase the rate of officially recorded crime. Improvements in computer technology and increases in the number of police personnel devoted to the collection of crime data, for example, have both contributed to “artificial” increases in crime within some jurisdictions (Siegel and McCormick 2006). Other methodological issues related to UCR statistics include:

In sum, official UCR crime statistics have specific strengths, as well as specific weaknesses. As a result of these weaknesses, criminologists often attempt to expand their analysis of crime by considering unofficial statistics — usually collected through self-report (victimization) surveys. We turn to these data in the next section.

Measuring Crime with Unofficial Data


The limits of official (UCR) statistics have caused social scientists to seek alternative methods for documenting the extent and nature of criminal behaviour in Canada. General population surveys are by far the most widespread strategy for collecting “unofficial” crime data. Two types of surveys have been developed: 1) self-report surveys — in which respondents are asked if they themselves have engaged in specific criminal behaviours over a given time period; and 2) victimization surveys — in which respondents are asked if they have experienced specific types of criminal victimization over the past year.8 In general, surveys uncover much more criminal activity than official police statistics do. For example, the 2004 Canadian General Social Survey (Gannon and Mihorean, 2005, discussed in detail below) produced an unofficial crime rate of approximately 28,000 per 100,000. By contrast, the 2004 crime rate produced by official UCR statistics was only 8,951 per 100,000 (see Table 1 in Silver 2007). The huge discrepancy between these two rates of crime can be explained by the fact that most crimes are never “discovered” by the police and thus recorded in UCR tabulations. Indeed, according to the results of the 2004 General Social Survey, only a third of all victimization incidents (34 per cent) are reported to the police.

The General Social Survey

The General Social Survey (GSS) is the only national victimization survey conducted in Canada. It is also the largest (in terms of sample size) and most methodologically sophisticated (in terms of sampling strategy and survey construction). The victimization cycle of the GSS is a telephone survey that is administered to a representative sample of Canadians aged 15 years and older. The most recent cycle of the victimization survey was administered in 2004. This was the fourth time that the survey was administered, with previous cycles in 1988, 1993 and 1999 (see Gannon and Mihorean 2005).

To select the sample, Random Digit Dialing technology is used to randomly select households in the ten provinces and the territories. Thereafter, one person from each household is randomly selected to complete the survey. The sample in 1988 and 1993 consisted of only 10,000 households. In 1999, the sample consisted of 26,000 households. In 2004, the final sample included 24,000 households.

Since the survey is conducted by phone, it automatically excludes people who do not have a home phone and people who are living within institutions. In the most recent iteration of the survey, this amounted to approximately four per cent of the Canadian population. Statistics Canada maintains that the loss of this “unreachable” population should not significantly impact national crime estimates.9 However, it should be noted that previous research has demonstrated that homeless/transient populations (i.e., those who are the least likely to have a phone) and those living in institutional facilities (including prisons) tend to have significantly higher rates of victimization than others have. Thus, though significantly higher than UCR statistics, the crime estimates provided by the GSS may still be conservative.

The results of the 2004 GSS indicate that 28 per cent of the Canadian population 15 years of age or older experienced at least one criminal victimization in the previous 12 months. The findings also indicate that eight per cent had experienced a physical assault in the past year, two per cent had experienced a sexual assault and one per cent had experienced a robbery. In sum, 10.6 per cent of the population had experienced one or more violent victimizations in the previous 12 months. This was down slightly (minus five per cent) from the rate of violent victimization recorded by the GSS in 1999.

Graph: Figure 15: Rate of Self-Reported Violent Victimization (per 1,000) by Province 2004 General Social Survey
Figure 15: Rate of Self-Reported Violent Victimization (per 1,000) by Province 2004 General Social Survey

Figure 15 presents the GSS violent victimization rates by province. In general, the data pattern is consistent with the UCR statistics presented above (see Figure 1). Over all, according to this national victimization survey, violent crime appears to be more prevalent in Western Canada than it is in Eastern Canada. One exception to this general rule is Nova Scotia, which recorded the second-highest rate of violent victimization in the country. Interestingly, while Manitoba and Saskatchewan have the highest rate of violent crime according to official UCR estimates, Alberta leads the way with respect to self-reported victimization. Similarly, while Ontario ranks eighth with respect to officially recorded violent crime, it rises to sixth when estimates are based on GSS data. Finally, according to UCR statistics, British Columbia’s official rate of violent crime (1,218 per 100,000) is 61 per cent higher than Ontario’s rate (756 per 100,000). However, according to the 2004 GSS, B.C.’s violent victimization rate (10,800 per 100,000) is actually slightly lower than Ontario’s rate (11,200 per 100,000).10 As discussed above, these regional discrepancies could be the result of regional differences in behaviour — or they could stem from divergent police practices.

Graph: Figure 16: Rate of Self-Reported Violent Victimization (per 1,000) by City 2004 General Social Survey
Figure 16: Rate of Self-Reported Violent Victimization (per 1,000) by City 2004 General Social Survey

Figure 16 presents GSS violent victimization rates for 13 major Canadian urban areas. According to 2004 survey results, Halifax, Nova Scotia (229 per 100,000) had the highest rate of violent victimization in Canada, followed by six Western cities (Edmonton, Regina, Calgary, Winnipeg, Victoria and Saskatoon). However, it should be noted that Ottawa’s violent victimization rate (143 per 100,000) was almost as high as Saskatoon’s (146 per 100,000). However, both Toronto and Hamilton are close to the national average (106 per 100,000). Interestingly, while Vancouver’s official (UCR) rate of violent crime is significantly higher than Toronto’s (see Figure 6 above), according to the 2004 GSS the two cities have identical rates of violent victimization. Finally, consistent with UCR statistics, the GSS found that Quebec City and Montreal have relatively low rates of violent crime. Over all, the basic crime pattern produced by the 2004 General Social Survey is largely consistent with the UCR police statistics compiled annually by the Canadian Centre for Justice Statistics. In general, violence is much more prevalent in Western Canada than it is in Eastern Canada (with the exception of Nova Scotia). Ontario and Ontario’s urban centres, on the other hand, tend to lie somewhere in the middle: less violent than Western Canada (and Nova Scotia), but somewhat more violent than Quebec.

As with police official statistics, the results of the 2004 GSS suggest that violence is highly concentrated among youthful populations (see Figure 17). Indeed, the violent victimization rate for respondents 15 to 24 years of age (226 per 100,000) is two times greater than the rate for 35-to 44-year-olds (115 per 100,000) and more than four times greater than those 55 years of age or older (45 per 100,000).

Graph: Figure 17: Rate of Self-Reported Violent Victimization (per 1,000) by Age Group 2004 General Social Survey
Figure 17: Rate of Self-Reported Violent Victimization (per 1,000) by Age Group 2004 General Social Survey

Respondents to the 2004 GSS were also asked if they could identify the ages of the offenders involved in violent victimization incidents. Only two per cent of respondents claimed that they could not estimate the age of the offenders involved in violence-related incidents (this percentage was much higher for property crimes). The results further demonstrate the robustness of the negative age-crime correlation. Two thirds of all offenders were identified as being less than 34 years of age. By contrast, only five per cent were 55 years of age or older. As with UCR data, violent offenders are particularly well represented among young adults. Indeed, while only 13 per cent of all offenders fall into the legally defined “young offender” category (12 to 17 years of age), 50 per cent of offenders were identified as being young adults between 18 and 34 years of age.

Graph: Figure 18: Estimated Age of the Offender(s) Involved in Violent Victimizations 2004 General Social Survey
Figure 18: Estimated Age of the Offender(s) Involved in Violent Victimizations 2004 General Social Survey

Other Youth Surveys

By all accounts, the General Social Survey (GSS) is the most ambitious victimization survey in Canada. However, over past decade, a number of other surveys have examined the nature and extent of youth violence in Ontario. Most of these studies have examined large samples of high school students. This section will focus on the results of ten of these studies to further illustrate the degree of violent behaviour among young people in Canada. The studies to be discussed include:

  1. The Ontario Student Drug Use Survey (OSDUS): This Ontario-wide survey has been conducted every two years since 1977 (Paglia et al 2003; CAMH 2006). The survey is sponsored and administered by the Centre for Addiction and Mental Health (CAMH, formerly known as the Addiction Research Foundation). Respondents are drawn from elementary schools (grades 7 and 8) and high schools (grades 9 to 12) from across the province. Each survey has involved a large, representative sample of Ontario students. For example, the 2005 survey involved a random sample of 7,726 respondents representing 42 different school boards and 137 different schools. Although most survey questions focus on student drug and alcohol use, questions about youth violence have been included since 1983.
  2. The International Youth Survey (IYS): This survey was administered to students in 30 different nations in 2006. The Canadian version, conducted by Statistics Canada, involved a random sample of 3,200 students (in grades 7 through 9) from the Toronto District School Board (see Savoie 2007).
  3. The 2006 Toronto District School Board Student Census: This study, completed in 2006, involved an extensive survey of approximately 105,000 students (grades 7 through 12) from the Toronto District School Board. The final sample included approximately 92 per cent of all students in grades 7 and 8 and 81 per cent of all students in grades 9 through 12. The survey included several questions on school safety and violence (see Yau and O’Reilly 2007).
  4. The 2000 Toronto Youth Crime and Victimization Survey (YCVS): This survey, completed in 2000, involved a random sample of 3,393 Toronto high school students (from both the Catholic and public school boards) and a random sample of 396 homeless street youth (response rate = 82 per cent). This survey investigated both self-reported victimization experiences and self-reported delinquent behaviour (see Tanner and Wortley 2002).
  5. The Drugs, Alcohol and Violence International Survey (DAVI): This study, completed in 2003, is a joint US-Canada investigation into youth drug use and violence. The study involved a survey of three samples of male youth 14 to 17 years of age: students, dropouts and young offenders being held in secure custody facilities. The Canadian version of the project collected data from samples in both Montreal and Toronto. The Canadian student sample consisted of 904 respondents — 456 Toronto high school students (from eight different schools) and 456 high school students from Montreal (from eight different schools). The dropout sample consisted of 218 respondents (116 from Toronto and 102 from Montreal). Finally, the young offender sample consisted of 278 youth in secure custody facilities (132 from southern Ontario and 146 from the Montreal region). The survey included many questions about weapons use and violent behaviour. As part of this project, similar surveys were conducted in both Philadelphia and Amsterdam (see Erickson and Butters 2006).
  6. School Community Safety Advisory Panel Surveys: The School Community Safety Advisory Panel (led by Julian Falconer) was formed by the Toronto District School Board in June 2007 to investigate issues of school safety in the wake of the tragic shooting death of Jordan Manners. As part of its investigation, the panel conducted victimization surveys at two schools located within an economically disadvantaged neighbourhood in northwest Toronto (see School Community Safety Advisory Panel 2008). The first survey was conducted in June 2007. The second survey was administered in October 2007. A total of 1,293 students participated in this research project (423 students at C.W. Jefferys Collegiate and 870 at Westview Secondary School).
  7. The 2001/2002 Health Behaviour in School-Aged Children Survey (HBSC): In 2001/2002, the World Health Organization (WHO) conducted a massive survey of 162,306 students from 35 countries. As part of this project, the WHO questionnaire was administered to a random sample of 4,361 Canadian youth. Several questions asked respondents to report on their experiences with both bullying and fighting (see Craig and Harel 2004).
  8. Canada’s National Longitudinal Survey on Children and Youth (NLSCY): This is a joint project of Statistics Canada and Human Resources Development Canada. It is a national, longitudinal survey of a group of children who will be studied every two years until they reach the age of 25 years (sample size = 22,831). The data on delinquency reported below are based on self-report forms filled out by the children themselves and returned to Statistics Canada (see Sprott and Doob 2008; Sprott et al 2001).
  9. The Ontario Student Sexual Harassment Survey: This project was conducted by researchers from the Centre for Addiction and Mental Health (CAMH). The survey, completed in 2007, involved a random sample of 1,800 students from 23 high schools in Ontario. A 75-minute-long questionnaire was first administered to the student respondents at the beginning of grade 9. The same respondents were re-interviewed two years later as they started grade 11. Most of the questions focused on issues of sexual harassment (Wolfe and Chiodo 2008).

In the following sections, we provide basic estimates of youth violence that have been derived from the nine studies outlined above. The objective of this exercise was to illustrate that various forms of violence are quite common in the lives of young people. We should note, however, that it is very difficult to compare the results of different studies. First of all, question wording varies dramatically from study to study, making direct comparisons impossible. Secondly, different studies employ different time frames in their analysis. For example, some studies ask about exposure to violence over the past few months, while other studies ask about exposure over the past year or over the course of a lifetime. Unfortunately, a common methodological benchmark for studying youth violence in Ontario has not yet been developed.

Bullying

Bullying has been defined as a form of abuse at the hands of peers. As such, it can take many forms: from threats, to physical assaults, to insults, to social exclusion (see Hunt 2007; Pepler et al 2006). As described by Craig, Pepler and Blais (2007), bullying represents a pattern of repeated aggression in which there is a power differential. Children and youth who bully always have more power than their victims. Their power might stem from greater physical size or strength or from social advantage (higher socio-economic position, higher status in a peer group, knowledge of a victim’s personal vulnerabilities, etc.) Research suggests that concern over bullying is warranted. Children who engage in bullying, for example, are at risk of developing long-term problems with aggression, anti-social behaviour and substance abuse. The victims of bullying, on the other hand, are at risk of developing serious mental health problems including depression, anxiety and somatic complaints (see Olweus 1993; Farrington 1993; Craig, Pepler and Blais 2007). Unfortunately, there is also considerable evidence to suggest that bullying is quite widespread among Canadian children and youth.

Physical Threats

Survey research also indicates that physical threats — both with and without weapons — are rather common behaviours among Canadian and Ontario youth. The following findings illustrate this point:

Physical Assaults

Surveys attempt to measure the extent of physical assault in a given jurisdiction by asking two types of questions: 1) how often the respondent has been assaulted (punched, kicked, slapped, etc.) over a given time period; or 2) how often the respondent has hit someone else or has been in a fight. Once again, survey results suggest that common assault and/or fighting are quite widespread among Canadian youth. For example:

Robbery/Extortion

Although much less prevalent than bullying, threats, or physical assaults, research suggests that a significant proportion of Ontario youth will experience robbery or extortion at some point in their lives:

Weapons Use

Over the past decade, the use of weapons among young people has become a major social issue in Canada. The public is particularly concerned about an alleged increase in the use of firearms. Survey results suggest that guns and other weapons are indeed used by a small but significant population of young people in Canada. Research findings also suggest that the use of weapons may be particularly high among youth from economically disadvantaged communities:

Sexual Assault and Harassment

In general, male youth are much more likely than female youth are to be both the victims and perpetrators of violence. The exceptions are sexual assault and sexual harassment. Although the vast majority of sex offenders continue to be male, the majority of victims are female. A number of recent youth surveys confirm that that a relatively high proportion of Ontario females will suffer from some form of sexual victimization at some point in their lives:

Results from the National Longitudinal Survey of Children and Youth (NLSCY)

In a recent article, Sprott and Doob (2008) examine data from the fourth cycle of the National Longitudinal Survey of Children and Youth (NLSCY). The data are weighted to produce a representative sample of Canadian adolescents 12 to 17 years of age. The NLSCY is an important study because it is the only self-report survey of Canadian youth that is conducted across Canada and thus permits provincial comparisons. Cycle Four of the NLSCY asked respondents about whether they had engaged in any of nine serious violent activities over the past 12 months: attack so severe that the victim required medical attention; assault with a weapon; carrying a knife; carrying a gun; carrying another weapon like a stick or a club; robbery; minor sexual assault (uninvited sexual touching); major sexual assault (forced someone to have sex against their will); and arson.

The results of the NLSCY indicate that 19.5 per cent of Ontario youth (aged 12 to 17) had engaged in at least one seriously violent behaviour in the past 12 months (Sprott and Doob 2008).15 According to the NLSCY, Ontario’s self-reported rate of violent behaviour is only slightly higher than the rate of serious violence reported by youth in Quebec (17.5 per cent), Saskatchewan (18.5 per cent) and British Columbia (17.7 per cent). However, the survey did find higher rates of violent offending in all other provinces, including those in the Atlantic region (20.0 per cent), Alberta (22.5 per cent) and Manitoba (26.4 per cent). Interestingly, provincial differences documented by the NLSCY are quite small. In fact, the provincial differences documented by the NLSCY are much smaller than those produced by statistics on the official processing of young offenders by province (see full discussion in Sprott and Doob 2008). Nonetheless, the results of this survey further illustrate that a significant proportion of Canadian youth — including those who reside in Ontario — have recently engaged in seriously violent behaviour.

Reporting Violence to the Police

As discussed above, the rates of violent victimization documented by surveys are typically much higher than the rates produced by official police statistics. This finding can be explained by the fact that many violent incidents are never reported to the police. The 2004 General Social Survey (GSS), for example, found that only 33 per cent of the violent incidents documented by the study were actually reported to the police. The 2004 GSS also indicated that police reporting rates vary dramatically by the type of crime. Indeed, 46 per cent of robbery victims and 39 per cent of assault victims reported their experiences to the police, compared with only eight per cent of sexual assault victims. Further analysis reveals that young people are particularly reluctant to report their own victimization experiences to the police. For example, only 24 per cent of victims aged 15 to 24 years decided to report their victimization to the police (see Gannon and Mihorean 2005).

A number of other studies confirm that young people rarely report their violent victimization experiences to the police:

Ontario at the Crossroads: Disturbing Trends in Youth Violence


The crime statistics presented above provide reason for optimism — as well as cause for concern. There is no doubt that Ontario has its share of violent crime. We are not a violence-free society. At the same time, however, official (UCR) crime statistics suggest that, by world standards, Ontario is a relatively safe place to live. Indeed, Ontario’s official rate of violent crime is significantly lower than the rates found in most other Canadian provinces. Similarly, the rates of violent crime reported for Ontario’s major urban areas tend to be significantly lower than the rates found in major American and European cities. There is also very little evidence to suggest that violent crime in Ontario is increasing. Indeed, despite public opinion to the contrary, violent crime in Ontario has actually decreased since the mid-1970s. Furthermore, Ontario’s overall rate of violent crime – including homicide — has been relatively stable since the mid 1990s.

These optimistic figures, while somewhat comforting, should not cause complacency. As a matter of fact, there are a number of disturbing trends with respect to youth violence in Ontario that deserve special attention:

In summary, by both national and international standards, Ontario remains a relatively safe place to live. Toronto, the province’s largest city, still has a remarkably low rate of violent crime — especially when compared with other large American and European cities. However, as the data above suggest, there is reason for concern. Fear of crime and violence is a growing problem in this province. There is also evidence to suggest that official crime statistics may dramatically underestimate the true level of violent victimization among young people. Finally, though overall crime rates have remained stable, severe violence is apparently becoming more and more concentrated among socially disadvantaged minority youth. Most disturbingly, recent data suggest that this general pattern of violence may become more entrenched if current economic trends continue.

Through numerous community consultations and an extensive review of the academic literature, we have come to the conclusion that the “roots of youth violence” are often found in poor, socially deprived neighbourhoods. Deprived neighbourhoods often experience a series of interrelated problems, including: family dysfunction (including high rates of teenage births and single-mother households); high youth unemployment; low household incomes; youth alienation and hopelessness; racism; poor levels of educational attainment; peer delinquency; mental health issues; and poor physical health outcomes. These factors often contribute, directly or indirectly, to increased rates of youth violence. Unfortunately, there is considerable evidence to suggest that these types of neighbourhoods are actually becoming more and more prevalent within Ontario. Take Toronto as a case in point. A series of reports commissioned by the United Way of Greater Toronto (see United Way of Greater Toronto 2005) have documented the following trends:

These data should be viewed as disturbing. It appears, unfortunately, that the gap between the rich and the poor, the haves and the have-nots, is widening. Ontario, it seems, is currently at risk of developing the types of permanent underclass communities (often referred to as ghettos) that have marked the history of urban development in the United States. The warning must be sounded: if such deprived neighbourhoods become entrenched, it is very likely that much higher rates of violent crime will follow. Clearly Ontario is at a crossroads. What changes will the next 20 years produce? Will Ontario — and Ontario’s cities — continue to be safe? Or are we headed towards becoming a more economically divided, more violent society? The policy decisions we make over the next five years may seal our fate.

References


Archer, Dane and Rosemary Gartner. (1984). Violence and Crime in Cross-National Perspective. New Haven: Yale University Press.

AuCoin, Kathy (2005). Children and youth as victims of violent crime. Juristat 25 (1). Ottawa: Canadian Centre for Justice Statistics.

Bailey, William and Ruth Peterson. (1999). Capital punishment, homicide and deterrence: An assessment of the evidence. In M. Smith, Dwayne Zahn and Margaret Zahn (Eds.), Studying and Preventing Homicide: Issues and Challenges (pp. 223–245). Thousand Oaks CA: Sage.

Besserer, S. and C. Trainor. (2000). Criminal victimization in Canada, 1999. Juristat 20 (10). Ottawa: Canadian Centre for Justice Statistics.

Canadian Centre for Addiction and Mental Health (CAMH). (2004). One in ten Ontario students have carried a weapon. CAMH Population Studies eBulletin 5 (Mar/Apr).

Canadian Centre for Addiction and Mental Health (CAMH). (2006). One in fifty Ontario students reports carrying a gun. CAMH Population Studies eBulletin 7 (Jan/Feb).

Carrington, Peter. (2001). Population Aging and Crime in Canada, 2000–2041. Canadian Journal of Criminology and Criminal Justice, 43 (2), 331–356.

Chettleburgh, Michael. (2007). Young Thugs: Inside the Dangerous World of Canadian Street Gangs. Toronto: Harper-Collins.

Craig, Wendy and Yossi Harel. (2004). Bullying, physical fightings and victimization. In Candice Currie, Chris Roberts, Anthony Morgan, Rebecca Smith, Wolfgang Setterobulte, Oddrun Samdal and Vivian Rasmussen (Eds.). Young People’s Health in Context: Health Behaviour in School-Aged Children (HBSC) Study: International Report from the 2001/2002 Survey. (pp. 133– 144). Geneva: World Health Organization.

Craig, Wendy, Debra Pepler and Julie Blais. (2007). Responding to bullying: What works. School Psychology International, 28 (4), 465–384.

Craig, Wendy and Debra Pepler. (1997). Naturalistic Observations of Bullying and Victimization on the Playground. Toronto: LaMarsh Centre for Research on Violence and Conflict Resolution.

Doob, Anthony and Carla Cesaroni. (2004). Responding to Youth Crime in Canada. Toronto: University of Toronto Press.

Doob, Anthony, Jane Sprott and Cheryl Webster. (2008). Understanding Youth Crime: The Impact of Law Enforcement Approaches on the Incidence of Violent Crime Involving Youth and Matters Related to Understanding the Implications of These Findings. Toronto: Report Prepared for the Provincial Review of the Roots of Youth Violence.

Erickson, Patricia and Jennifer Butters. (2006). Final Report: Youth, Weapons and Violence in Toronto and Montreal. Ottawa: Public Safety and Emergency Preparedness Canada.

Farrington, D. P. (1993) Understanding and preventing bullying. In M. Tonry (Ed.) Crime and Justice, 17 (pp. 381–458). Chicago: University of Chicago Press.

Field, Simon. (1999). Trends in Crime Revisited. London: Home Office.

Fitzgerald, Robin. (2003). An Examination of Sex Differences in Delinquency. Crime and Justice Research Paper Series. Ottawa: Canadian Centre for Justice Statistics.

Gannon, Maire. (2005). General Social Survey on Victimization, Cycle 18: An Overview of Findings. Ottawa: Statistics Canada.

Gannon, Maire and Karen Mihorean. (2005). Criminal victimization in Canada, 2004. Juristat 25 (7). Ottawa: Canadian Centre for Justice Statistics.

Gartner, Rosemary and Sara Thompson. (2004). Trends in homicide in Toronto. In Bruce Kidd and Jim Phillips (Eds.) From Enforcement to Prevention to Civic Engagement: Research on Community Safety. (pp. 28–42). Toronto: Centre of Criminology, University of Toronto.

Hunt, Caroline. (2007). The effect of an education program on attitudes and beliefs about bullying and bullying behaviour in junior secondary school students. Child and Adolescent Mental Health 12 (1), 21–26.

Li, Geoffrey. (2007). Homicide in Canada, 2006. Juristat, 27 (8). Ottawa: Canadian Centre for Justice Statistics.

MacQueen, Ken. (2008, March 12). “The most dangerous cities in Canada.” Maclean’s 121 (11), 38–42.

Olweus, D. (1991). Bully/victim problems among school-children: Basic facts and effects of a school based intervention program. In Debra Pepler and K. Rubin (Eds.), Development and Treatment of Childhood Aggression (pp. 409–445). Hillsborough: Erlbaum.

Olweus, D. (1993). Bullying at School: What We Know and What We Can Do. Oxford: Blackwell.

Paglia, Angela and Edward Adlaf. (2003). Secular trends in self-reported violent activity among Ontario students: 1983–2001. Canadian Journal of Public Health 94 (3), 212–217.

Paetsch, Joanne and Lorne Bertrand. (1999). Victimization and delinquency among Canadian youth. Adolescence 34 (Summer), 351–367.

Pepler, Debra, Wendy Craig, Jennifer Connolly, Amy Yule, Loren McMaster and Depeng Jiang. (2006). A developmental perspective on bullying. Aggressive Behaviour 32, 376–384.

Savoie, Josee. (2007). Youth self-reported delinquency, Toronto, 2006. Juristat, 27 (6). Ottawa: Canadian Centre for Justice Statistics.

Savolainen, Jukka. (2000). Relative cohort size and age-specific arrest rates: A conditional interpretation of the Easterlin effect. Criminology 38, 117–136.

School Community Safety Advisory Panel. (2008). The Road to Health: A Final Report on School Safety. Toronto: Toronto District School Board.

Siegel, Larry and Chris McCormick. (2006). Criminology in Canada (3rd ed.). Toronto: Thomson.

Silver, Warren. (2007). Crime statistics in Canada, 2006. Juristat, 27 (5). Ottawa: Canadian Centre for Justice Statistics.

Sprott, Jane and Anthony Doob. (2008, forthcoming). Research note: Youth crime rates and the youth justice system. Canadian Journal of Criminology and Criminal Justice.

Sprott, Jane, Anthony Doob and Jennifer Jenkins. (2001). Problem behaviour and delinquency in children and youth. Juristat, 21 (4). Ottawa: Canadian Centre for Justice Statistics.

Statistics Canada. (2008, February 20). Firearms and violent crimes. The Daily.

Tanner, Julian and Scot Wortley. (2002). The Toronto Youth Crime and Victimization Survey: Overview Report. Toronto: Centre of Criminology, University of Toronto.

United Way of Greater Toronto. (2005). Poverty by Postal Code: The Geography of Neighbourhood Poverty – 1981–2001. Toronto: United Way of Greater Toronto and the Canadian Council of Social Development.

Yau, Maria and Janet O’Reilly. (2007). 2006 Student Census, Grades 7–12: System Overview. Toronto: Toronto District School Board.

Wallace, Marnie. (2003). Crime statistics in Canada, 2003. Juristat, 24 (6). Ottawa: Canadian Centre for Justice Statistics.

Wolfe, David and Debbie Chiodo. (2008). Sexual harassment and related behaviours reported among youth from Grade 9 through 11. Canadian Centre for Addiction and Mental Health (CAMH).

Wortley, Scot. (2007). Perceptions of Crime and Criminal Justice Among Toronto Residents: Results from a General Population Survey. Toronto: Centre of Criminology, University of Toronto.


1 It is interesting to note that some property crimes — including car theft — are more likely to be reported to the police than serious violence. Members of the public are often motivated to report serious property crimes for insurance purposes. Insurance claims cannot be made without a police report.

2 It should be noted that robbery is an offence that is both a property crime and a violent crime. However, according to UCR protocols, robbery is only counted under the violent crime category.

3 This figure is consistent with other data that suggest that, in an average year, the Canadian police identify a suspect in only 30–40 per cent of reported criminal incidents. This statistic is often referred to as the clearance rate (see Carrington 2001).

4 Age-crime data from the UCR database typically have to be purchased from Statistics Canada through its “special tabulation” service.

5 Unfortunately, we could not locate age-specific UCR data for Ontario.

6 Statistics Canada’s violent crime rate is a composite measure that includes homicide, manslaughter, attempted murder, common (level one) assault, assault with a weapon, aggravated assault, sexual assault and robbery.

7 We calculated age-specific homicide rates by utilizing 2006 homicide counts from the Canadian Centre for Justice Statistics (Li 2007; http://www40.statcan.ca/101/est101/legal10b.htm) along with age-specific population estimates from Statistics Canada (http://www40.statcan.ca/101/est01/demo10a.htm?sdi=age).

8 In rare cases, surveys try to capture both self-reported criminal activity and personal victimization experiences (see Tanner and Wortley 2002).

9 Gannon and Mihorean 2005: 19.

10 The GSS reports victimization rates per 1,000, while the UCR reports crime rates per 100,000. In order to make the figures more comparable, we multiplied the GSS rates by 100.

11 A definition of bullying was provided to respondents before they were asked the two questions about bullying. The definition stated that: “We say that a student is being bullied when another student, or group of students, says or does nasty and unpleasant things to him or her. It is bullying when a student is teased repeatedly in a way he or she doesn’t like, or when (he or she is) deliberately left out of things. But it is not bullying when two students of about the same strength quarrel or fight. It is also not bullying when the teasing is done in a friendly and playful way““ (Craig and Harel 2004: 133).

12 Sexual harassment was defined as unwanted sexual comments or suggestions that upset the respondent.

13 Minor sexual assault was defined as unwanted sexual touching or groping.

14 Major sexual assault was defined as being forced into sexual activity against your will.

15 Unfortunately, the authors only released statistics with respect to a composite scale of violent behaviour. The NLSCY has yet to release information on the percentage of youth who have engaged in specific types of violence (i.e., major sexual assault, assault with a weapon, etc.)

16 Student were identified as criminal gang members if they: a) reported current membership in a gang; and 2) reported that they engaged in criminal activity (drug dealing, fighting, theft, robbery, etc.) as part of this gang.

Contents

Volume 1. Findings, Analysis and Conclusions

Volume 2. Executive Summary

Volume 3. Community Perspectives Report

Volume 4. Research Papers

Volume 5. Literature Reviews