ARCHIVE - Final Report: "Youth, Weapons and Violence in Toronto and Montreal"

Archived Content

Information identified as archived is provided for reference, research or record-keeping purposes. It is not subject to the Government of Canada Web Standards and has not been altered or updated since it was archived. Please contact us to request a format other than those available.

Archived

This page has been archived on the Web.

March 31, 2006

A report prepared for Public Safety and Emergency Preparedness, Canada, under the direction of Project Manager, Rebecca Jesseman.

Dr. Patricia G. Erickson
Senior Scientist, Centre for Addiction and Mental Health
Professor of Sociology and Criminology, University of Toronto
Pat_Erickson@camh.net

Dr. Jennifer E. Butters
Director of Research, Centre for Urban Health Initiatives, University of Toronto
Jenn.Butters@utoronto.ca

The authors wish to acknowledge the following collaborators: Dr. Edward Adlaf, Centre for Addiction and Mental Health, Dr. Serge Brochu, University of Montreal and Dr. Marie-Marthe Cousineau, University of Montreal.

Table of Contents

Executive Summary

Objectives

This report presents research that contributes to knowledge about the inter-relationships of youth, guns and gang violence in Toronto and Montreal.

Method

The data base consists of three samples of male youth aged 14 to 17 years – students, dropouts and offenders in secure custody – recruited using the same questionnaires and data collection techniques in both cities. Between 2000-2003, a representative sample of students was surveyed in the classroom and dropouts and offenders were interviewed in private. A comparative analysis of the two cities and the three separate samples was conducted. This report presents both descriptive and multivariate findings.

Results

Several issues of concern were investigated. Results are complex and show variations between sites and samples, though some consistent patterns were displayed. While the prevalence of gang fighting was higher in Montreal than Toronto, in both cities it was a strong predictor of gun violence among both students and detainees. Selling drugs and specifically selling cocaine/crack significantly increased the odds of committing gun violence against others for Toronto detainees but not for those in Montreal. For dropouts, selling drugs was also positively associated with gun violence against others, but this relationship only reached significance in Toronto. Youth reported that obtaining a gun could be done in less time in Toronto than Montreal, reflecting an overall higher rate of gun exposure on various measures in that city.

Limitations

This report does not report trend data, since no prior Canadian research contains the detail and depth of questions on youth and firearms for base line purposes. The focus on adolescents, 14-17 years old, reflects the period when serious interpersonal violence increases and peaks; however, the data cannot be generalized to older youth, or to those in other cities and parts of Canada.

Recommendations

Adolescent males who carried and used guns were found, in differing degrees, among all three samples in both cities. For those who have taken this extreme step, enforcement and identification are essential for the protection of the public. To reverse the apparent trend of growing gun violence, and limit new gang recruits, a broader framework for large-scale prevention is also necessary. Different approaches should be considered for high-risk youth among the student, dropout and detainee groups. Preventive strategies in the form of realistic alternatives to gang membership and enhancement of school success are crucial for students. For dropouts, economic incentives to return to school and promote viable income sources instead of drug selling are a powerful tool. Delinquent youth require the most intense intervention, counseling and supervision in efforts to reverse a criminal trajectory into adult violent crime. Both short and long term steps are crucial in the overall plan for reducing gun-related violence in Canada.

"Youth, Weapons and Violence in Toronto and Montreal"

A report prepared for Public Safety and Emergency Preparedness, Canada, under the direction of Project Manager, Rebecca Jesseman (submitted March 31, 2006)

Patricia G. Erickson and Jennifer E. Butters, University of Toronto

I. Introduction

Recent attention by the media and public to the apparent increase in firearms related homicides in Toronto, Canada's largest and most ethnically diverse city, has tended to focus on "youth, guns and gangs" as the driving force behind this emerging trend. The search for explanations and potentially effective interventions has also revealed how little research is available in Canada to address these issues1. Our relatively low homicide rates and greater regulation of firearms access relative to our neighbour to the South may have contributed to a sense of complacency that "it couldn't happen here." In contrast, the high wave of concern about youth and guns in the USA has prompted a spate of research in the past 20 years. Consequently, we are in a position in Canada to perhaps over-draw from American data and conclusions, lacking our own evidence on the scope and nature of our Canadian problem. Fortuitously, a joint US-Canada project on youth, drugs and violence, the Drugs, Alcohol and Violence International [DAVI] study that began in 19992, can provide important and unique evidence on various dimensions of weapon involvement in Toronto. Additional resources made it possible to add Montreal to the study3, providing the first detailed data base in Canada to permit analyses of the inter-relationships among guns, gangs and youth violence.

This report will present the comparative findings for three groups of youth in Toronto and Montreal. Adolescents aged 14-17 were recruited – from schools, the community and correctional institutions – to reflect a continuum of embeddedness in conventional settings and a range of risk for delinquent activities. Because the students represent a random sample of youth in school, the majority of whom are not engaged in law-breaking (except of the most petty nature), they are considered the most normative group where the least violence would be expected. Youth who do not attend school regularly and are away from adult supervision are more likely to commit other offences, hence the dropouts are considered in the middle for potential violent involvements, and are followed by the highest risk group, the detainees who have been sentenced to custody as young offenders. All the data were collected in a 3 year time span, beginning in 2000 in Toronto and 2001 in Montreal, where the same methods and questionnaires (translated) were utilized to ensure comparability. Students were surveyed in classrooms and dropouts and detainees were interviewed in a one-to-one situation. While the study includes both boys and girls, this report will focus on the male samples only. A number of studies have consistently shown that about 3 to 4 times as many male youth as female youth report that they carry weapons, including firearms.4 Young men also tend to dominate when the most serious forms of violence are considered.5 However, a brief note on the separate analysis of girls and weapons (see footnote 1) will be included in the highlights section at the end of the report.

Other features of the original study design are advantageous to the task at hand. Since it was presented to youth as a study of "drugs and violence," no particular attention was drawn to an interest in either guns or gangs. This increases confidence in the validity of responses to these items, with no expectation that youth would feel directed to either under- or over- emphasize their involvement. The importance of placing violence in a context that included a variety of potential weapons was emphasized in our original proposal. However, having an American partner (University of Delaware) and a US funding source, ensured that a number of detailed questions specifically on firearms were included. Thus it is possible in this report to place these detailed questions on the carrying and use of guns within the framework of weapons use more generally. Similarly, since the original proposal was not framed as a study of gangs, different forms of gang membership were not elicited6; nevertheless, as part of our interest in delinquency, standard questions about "taking part in gang fights" were included. This implies the most serious and violent form of gang activity that is most relevant to linkage with firearms use.

This report is organized as follows. The methods and samples will be described in detail. Then the measurement of the variables of interest will be presented. Findings will be presented in two parts. First, the descriptive findings on a number of items depicting different types and locations of weapon carrying, the youths' experience of gun violence against themselves and against others, its relationship to gang fighting, a detailed account of the most violent incident they experienced with respect to its gun/gang relatedness, and attitudes to weapons, will be discussed for all three groups of youth in both cities. For the dropouts and detainees only, their accounts of how long it would take to get a gun, and the source, will be compared for Toronto and Montreal. Second, the multivariate analysis will present logistic regressions to control simultaneously for various factors that might explain violent outcomes with guns, and identify the most significant predictors. The report will conclude with highlights of the findings for cities and groups, and recommendations that follow in the areas of enforcement, prevention and policy.

II. Methods: Sample Selection

IIa. Student method & samples: Toronto & Montreal

The DAVI study for the male school sample is based on self-administered questionnaires of a total of 904, 9th- to 12th-graders from Toronto (8 schools; n=456 students) and Montreal (8 schools, n=448) between April 2001 and May 2003. In each site a stratified (region by income) two-stage (school, class) probability sample of students was employed. To ensure adequate variation in socio-economic status [SES] and region, in the first stage of selection the population of schools was stratified by region (Census Metropolitan Statistical Area [CMSA]-city vs. CMSA-outskirts) and SES (low vs. other), resulting in four strata. In each site, the standard CMSA was divided between the city metropolitan area and the surrounding area encompassed in the CMSA.

As well, based on census data, 'low income' schools were defined as those in which 20% or more of the area residents reported low income, versus other schools in areas reporting less than 20% low income. In each strata, two schools were randomly selected with probability proportional to size (PPS) sampling, resulting in 8 schools per site. In the second stage of selection, 1 class for each grade level G9 though G12 was selected within each school, resulting in 4 classes per school.

IIb. Dropout method & samples: Toronto & Montreal

A total of 218 male dropouts were interviewed in Toronto and Montreal. All respondents received $15 at the end of the interview.

Toronto Dropout Method The 116 Toronto dropouts were those who had left school for at least 30 consecutive days during the past 12 months. Personal interviews were conducted during a 29 month period between July 2000 and November 2002. The eventual sample was derived from 4 youth service agencies, 4 alternative school programs, 2 community centre agency drop-ins, direct call-in contacts from advertisements and street outreach. Interviews were conducted in private areas provided at the agency, typically an intake office or lunch room, or in private offices at the Centre for Addiction and Mental Health.

Montreal Dropout Method The 102 Montreal dropouts were 14 to 17 year olds derived from community agencies. Personal interviews were conducted between January and June 2003. The interviews were approximately one hour in length. Participants were recruited from 18 facilities throughout the Metropolitan area of Montreal. These facilities included: a) community organizations that directly help school drop-outs, b) community organizations that help street youth, c) organizations serving youth in need, d) youth centers, e) community organizations that help youth find jobs, f) and general drop-in centers. In each the coordinator of the facility was contacted and assisted with sample recruitment and interview facilitation.

IIc. Detainee method & samples: Toronto & Montreal

The Toronto and Montreal highest risk samples are based on personal interviews with 278 male adolescents aged 14 to 17 years from 2 samples (Toronto Detainees n=132 and; Montreal Detainees n=146).

Toronto Detainee Method The 132 Toronto Detainees were youth who were sentenced to secure custody, but not awaiting hearings, and held in one of ten facilities in Southern Ontario. Personal interviews were conducted during a 3-year period between January 2000 and December 2002. The 10 correctional facilities in Southern Ontario represent all secure juvenile facilities that retain youth for 24 hours or longer. While these facilities hold youth from all parts of the province, the sample was drawn only from those whose family home fell within the Toronto CMSA. Private, one-on-one interviews were conducted by research staff.

Montreal Detainee Method The 146 Montreal detainees were recruited from 10 custody facilities (5 serving the metropolitan area of Montreal, 2 the South Shore and 3 in Laval and the Laurentides). Personal one-on-one interviews were conducted in private in 2002 - 2003. Team coordinators contacted at each Youth Centre were instrumental in facilitating the interview process.

III. Measurement & Analysis

The following section outlines the variables used in the descriptive and multivariate analyses. Some slight variations will be noted for the different samples.

IIIa. Descriptive Analysis: Background Characteristics

Age: Respondents were asked to report their age. Due to the sampling selection criteria the age range was restricted to those 14 to 17 years (though a small number of respondents just outside those age ranges were included).

Race: Youth were asked "To which racial group would you say you mostly belong?" The response categories for the detainees and dropouts includes 'White', 'Black', 'Asian', and 'Other'. When presented with the questionnaire for the student survey, the participating school boards felt this was too restrictive for their population so additional categories were added: 'Aboriginal', 'Hispanic' and 'Mixed'.

Lives with both parents: The respondents were asked to indicate who lived in the same household as they did. Based on these responses a dichotomous indicator of intact family status was constructed so that individuals indicated they lived with both their mother and father were scored '1', while those living with only one parent were given a score of '0'.

School Suspension: Subjects were asked if they had "ever been suspended or expelled from school". Those who reported any lifetime suspension were given a score of '1' and those who had never been suspended or expelled were scored '0'.

Gang Fighting: A series of questions were asked regarding the respondent's history of participating in gang fights before they were 13 years of age and in the past 12 months. Youth who gave positive responses to participating in a gang fight before they were 13 years old were given a score of '1' and those who had not engaged in this behaviour before 13 were scored '0'.

Respondents were also asked "How often in the past 12 months have you taken part in a gang fight?" Responses to this item ranged from 0 to 100 times in the past year (depending on the sample). For the purpose of these analyses, this item was then recoded into a dichotomous variable where '0 times in the last 12 months' was scored '0' and responses of '1 or more times in the past 12 months.' were recoded to '1'. The new variable reflected any gang fighting in the past 12 months.

A new variable to reflect any lifetime participation in gang fighting was generated using these two items. Individuals who had participated in a gang fight before they were 13 years old, reported any gang fighting in the past year, or who reported both pre-13 and past 12 month gang fighting, were given a score of '1' and those who had no pre-13 or past year gang fighting were scored '0'.

Contact with the Justice System: This item varies according to the sample. Students in Toronto and Montreal were asked only if they had ever spent a night in jail or detention (yes=1; no= 0). Detainees were also asked to report the number of days they had spent in custody and the average number of days is provided in this report for descriptive purposes. Finally, the dropouts were asked whether they had ever been in jail or detention (1=yes; 0=no) and how many days they had spent in custody (the average number is cited in this report).

Drug Selling: These items also vary between the samples. In the student analyses the item 'past 12 month drug selling' was generated from the responses to two questions. Students were asked: "How often in the past12 months have you sold drugs to others" and "How often in the past 12 months have you received something of value in exchange for drugs." Those who responded often, sometimes or rarely were given a score of '1' and those who said never were scored '0' for both items. These two items were then summed and students who had either sold drugs and/or received something of value for drugs in the past year were given a score of '1'.

Detainees and dropouts were asked if they had ever sold or exchanged something of value for drugs (1=yes; 0=no). Those who indicated being involved in the drug market were then asked which drugs they sold. Previous literature clearly indicates that a higher degree of violence (including weapon related violence) is associated with certain types of drug markets. For the purpose of these analyses, a new item was created to reflect involvement in the cocaine/crack selling market. Individuals who said they had either sold cocaine or crack, or had sold both, were given a score of '1' and those who reported no involvement in either of these drug markets were scored '0'.

Weapon Carrying

Due to time constraints imposed by available class time, many of the weapons items were asked in a simple 'yes' 'no' format to students, collapsing the categories used in the detainee and dropout face-to-face interviews. These will be identified and summarized below.

A variety of questions are included in these analyses that pertain to the weapon carrying behaviour of others. These include the following:

  1. "In your school, how many students carry some sort of weapon?" with the response options of 'most', 'some', 'a few' and 'none'.
  2. "Do any of your friends carry weapons?" (1=yes; 0=no). This item was recoded in the detainee and dropout samples to reflect this dichotomy ('most', 'some' and 'a few' were coded 1 and 'none' coded 0).
  3. "Do you know anyone who has carried a gun to school"? (1=yes; 0=no). This item was recoded in the detainee and dropout samples to reflect this dichotomy (see above).

Four questions were also asked gauging the respondent's personal weapon carrying behaviour. These were:

  1. "If you carry a weapon, what weapon(s) do you carry?" Students, detainees and dropouts were asked to answer 'yes' or 'no' to the following weapon types: Mace, knife and gun. Detainees and dropouts were given two additional weapon types, martial arts weapons and club or stick.
  2. "Have you ever carried a weapon when you were not at school?"
  3. "Have you ever carried a weapon when you were at school?"
  4. "Have you ever carried a gun to school?"

For the purpose of these analyses, questions 2 through 4 were dichotomized for all three samples (1-yes; 0=no). Recodes were therefore applied to the detainee and dropout responses. These groups were asked to indicate 'how often' they had carried and weapon/gun outside of or in the school environment ('always', 'often', 'sometimes' or 'never'). These responses were recoded so that any amount of weapon/gun carrying was given a score of '1' and those who said never were scored '0'.

Finally, the three groups were asked a perceptual question: "in your school, how serious a problem is students bringing guns to school?" Three response options were given: 'very serious', 'somewhat serious' and 'not at all serious'.

Experience of Gun Related Violence

Students, detainees and dropouts were each asked questions about their lifetime and past 12 month involvement in gun related violence. These items captured their gun violence experiences against self and against others during these two time periods. Specifically, youth were asked "has someone ever threatened or tried to hurt you with a gun?" (yes=1; no=0). Those who indicated lifetime gun violence against themselves were then asked "how many times has someone threatened or tried to hurt you with a gun in the past 12 months?" Four categories of responses to this item are presented in this report: 0 times, 1 time, 2 times and 3+ times in the past 12 months.

Gun violence against others was measured the same way using the following two items: "Have you ever threatened or tried to hurt someone with a gun?" and "How many times have you threatened or tried to hurt someone with a gun in the past 12 months?"

Accounts of Most Violent Incident

All three of our samples were asked to provide specific details about the nature and context of the most violent event they had experienced. Characteristics pertaining to gang involvement, gun threats and injury, and drug use and dealing are included in this report.

Gang involvement:

Three gang related questions were asked. Youth were asked if gang members 'did' the violence (1=yes; 0=no), if gang members were the 'target' of the violence (1=yes; 0=no), and if they perceived the main cause of the event to be gang related (1=yes; 0=no).

Gun threats and injury:

The respondents were asked to describe the various types of violence that may have taken place during the event. In this report we include two: (a) someone threatened or tried to hurt someone with a gun and (b) someone got hurt with a knife, gun or other object. Although the latter item was not restricted to guns, information pertaining to the most serious and potentially lethal outcome of a violent incident is important to capture.

Drug use and dealing:

The youth were asked to think about the day of the event and indicate whether they had used any of the following psychoactive drugs: alcohol, marijuana, cocaine, crack, hallucinogens, amphetamines, and heroin. In the original questionnaire attempts were made to gauge the potential degree of intoxication at the time of the incident. To this end the respondents were given four response options: "did not use", "used but not high at the time [of the incident]", "used and was high at the time [of the incident]", and "used and was coming down [at the time of the incident]". For the descriptive purpose of this report, these items were recoded to reflect ('1') any use, regardless of perceived state of intoxication at the time of the incident and no use ('0'). Following this an overall dichotomous indicator of any psychoactive drug use on the day of the most violent event was created. Specifically, respondents reporting having used one to seven substances were given a score of '1' and no drug use on the day of the incident was coded '0'.

To estimate the potential association between the violent event and the illicit drug market, respondents were asked to indicate (yes or no) whether the event was in any way related to drug dealing activities.

Attitudes Towards Weapons

Two items regarding the respondent's attitudes toward why teenagers may carry weapons are included in the analyses. All three groups of respondents were asked to either agree or disagree with the following statements: "if you really want to hurt someone a weapon is the easiest way to do it" and "when you carry a weapon you get a lot more respect." These items were then recoded to reflect positive weapon attitudes (agree =1; disagree=0). In retrospect, for these items it would have been beneficial to ask the respondents specifically about gun carrying and their own attitudes and reasons for doing or not doing so.

IIIb. Dependent Variables

Lifetime gun related victimization and aggression are measured using the two lifetime items previously described in the "Experiences with gun related violence" section in this report (see above). For the student sample a single outcome variable is employed. Because reports of gun violence against self and towards others are minimal in this group, these actions have been combined to reflect a measure of any lifetime experience of gun related violence against oneself and/or towards others. Specifically, those students who report either or both forms of gun related violence are given a score of 1. Those who have not experienced any form of gun related violence are given a score of 0. For the dropout and detainee samples, where gun violence is reported more frequently, the two gun related outcomes of interest are kept separate for the purpose of these analyses. These are: (i) lifetime gun violence against self; and (ii) lifetime gun violence against others. For each measure, detainees and dropouts who have experienced violence are scored 1 and those who have not are scored 0.

IIIc. Binomial Logistic Regression Analysis

Logistic regression is useful for research questions where the object is to predict the presence or absence of an outcome based on a set of explanatory variables. It is suited to models where the dependent variable is dichotomous (two-categories). Logistic regression coefficients can be used to estimate odds ratios for each of the independent variables in the model as they predict the likelihood of the outcome of interest. Coefficients that are greater than 1 suggest an increase in the odds or likelihood of the dependent variable. Those coefficients less than 1 imply a reduction in the odds. In addition to the direction and magnitude of the coefficient (odds ratio), these analyses also take significance into consideration. A coefficient is produced for each independent (predictor) variable in the model; however, not all of these variables have a significant effect on the outcome of interest. In the following analyses significance levels (p) are denoted using the following symbols ranging from least to most significant: * p<.05; ** p<.01 and *** p<.001.

In this report the results of the logistic regression analysis are presented in odds ratios (ORs) that examine the predictive value of the demographic, custody, school suspension, gang fighting, drug dealing and weapon attitudes variables for the three gun related violence outcomes. For the purpose of the logistic regression analyses, age was recoded into a two category variable reflecting the younger and older age groups of the respondents (0= 14-15 year olds; 1= 16-17 year olds). In addition, because several categories comprise the race variable, the results are presented in comparison to those who self-identified as being 'White'.

IV. Sample Characteristics

IVa. Background Characteristics: Students

Differences between the two male student samples were evident in age and race composition (Appendix A, Table 1). The students were slightly older in Toronto (64.8% 16 – 17 years of age) and younger in Montreal (58.5% 14 - 15 years of age). In Montreal a majority of male students were White (75%), while in Toronto less than half were White and close to one-third were of Asian backgrounds. Students were similar in the proportion living with both parents (about 3/4).

Differences were also noted in the proportion of students who reported ever being suspended, participating in a gang fight (ever and before they were 13) and spending any nights in jail or detention. In each case, a greater number of students in Montreal reported these experiences. Conversely, a slightly higher percentage of Toronto students reported selling drugs in the past 12 months (15% vs. 10%).

IVb. Background Characteristics: Dropouts

In both Toronto and Montreal 77% of the dropout youth were 16 to 17 years of age (Appendix A, Table 2). In Montreal, two-thirds of the youth were White and 18% were Black. In Toronto, 43% of the dropouts were White and roughly one-third were Black. Approximately 30% of Montreal dropouts reported living with both parents while only 12% reported the same in Toronto.

The two groups of male dropouts differed on several items. Although the dropout sample represents those youth who had not been in school for 30 consecutive days in the past year, a greater proportion of the youth in Toronto indicated having ever been expelled or suspended from school. Reports of gang fighting were more prevalent among the Montreal dropouts. Specifically, 60% indicated ever taking part in a gang fight (38% in Toronto), 38% had done so before they were 13 (22% in Toronto) and three times as many had participated in 3 or more gang fights in the past year (29% in Montreal compared to 9.5% in Toronto).

A greater proportion of Toronto dropouts reported ever having spent a night in jail or detention than those in Montreal (71% vs. 21%), but the average number of days spent in jail or detention was far greater for the dropouts in Montreal (189 average days vs. 69 average days). Finally, Toronto dropouts were more likely to report having ever sold drugs (69% vs. 51%) and to have sold cocaine and/or crack (21% vs. 10%) than those in Montreal.

IVc. Background Characteristics: Detainees

As illustrated on Appendix A, Table 3, differences were observed in age and racial composition for the samples. While in both Toronto and Montreal a greater proportion of youth were 16 to 17 years of age, this was particularly evident in Montreal where less than 10% of the male detainees were 14 – 15 years old. The Montreal sample is also predominantly White (71%), whereas in Toronto 55% of the male detainees were White and approximately one-quarter were Black. Similar proportions reported living with both parents (24% Toronto & 21% Montreal),

The samples of detainees were similar on a number of delinquency variables. Nearly all have been suspended or expelled from school at some time, and two-fifths reported taking part in a gang fight before they were 13 years of age (40% Toronto and 42% Montreal). In addition, for both samples more than a third (36% Toronto and 39% Montreal) reported participating in 3 or more gang fights in the past 12 months. The proportion of detainees who reported ever selling drugs, and selling cocaine, crack or both substances, is also very similar in both cities; over 80% reported ever selling drugs and approximately 40% indicated selling cocaine and/or crack.

Finally, a noticeable difference is evident in the average number of days spent in custody. Male detainees in Montreal reported dramatically longer periods of stay in youth justice facilities (502 average days vs. 117 average days in Toronto).

V. Descriptive Findings

The tables that accompany the descriptive results are presented in Appendices B through D. The following sections provide a general overview of the findings. This portion of the report is meant to offer a descriptive snap-shot of the observable trends in the data regarding weapons and gun related violence. Analyses of statistical significance are restricted to the multivariate analysis where the basis for making inferences is much stronger and more applicable.

Va. Students

Weapon Carrying Behaviour

Weapons in the school environment appeared to be more common in Toronto than Montreal (Appendix B, Table 4). Among the Toronto students, 77% indicated 'some' or 'a few' of the students in their school carried weapons in comparison to 56% in Montreal. One-fifth (22%) indicated knowing someone who had brought a gun to school in contrast to 7% in Montreal. It perhaps is not surprising then that one-third of the Toronto students perceived guns to be a 'very or somewhat serious' problem in their schools, while only 18% of the students in Montreal felt the same. Interestingly, students in Montreal were somewhat more likely to self-report carrying weapons, including a gun, into the school environment. The students in Montreal and Toronto were equally likely to indicate having ever carried a weapon when not in a school setting (40% in each city).

Finally, in both cities, the greatest proportion of students reported carrying knives (20% in Toronto and 18% in Montreal) over other choices of weapon. Guns were reported by 4.2% in Toronto and 2.8% in Montreal. These rates are similar to those of mace carrying (4% and 3.2%).

One critical issue for understanding weapon carrying behaviour pertains to the context in which youth feel the need to arm themselves. One potential contributing factor may be involvement in gang related activities. Table 5 in Appendix B presents a series of analyses that highlight the relationship between the proportion of students who indicated carrying weapons (and specifically guns) who also reported ever participating in a gang fight (though these numbers are small).

Among those who reported ever carrying a weapon when they were not at school, gang fighting was potentially more relevant for the students in Montreal. Although for both groups the majority of students reported no gang fight involvement, among those in Montreal, 44% reported ever participating in a gang fight while 37% of the Toronto students reported the same. Gang involvement among those who reported carrying a weapon in the school environment appeared to be slightly more pronounced. While it was still the case that half of this subset indicated no gang involvement, there was an almost 50-50 split in Montreal while in Toronto, two-fifths (43%) of the students who have carried a weapon to school also indicated gang fighting. Finally, among the very small subset of Toronto students who have carried a gun to school (N=4), gang fighting appeared to be minimal. Once again participation in gang fights was greater among the Montreal students where 44% of those who carried a gun to school (N=9) also reported ever participating in a gang fight compared to 25% in Toronto.

Attitudes toward weapons

Why do teenagers carry weapons? What are their attitudes towards weapons?

An important piece of the puzzle in beginning to understand why youth decide to carry weapons including guns is to gauge their attitudes toward them. Two potentially informative indicators are whether young people feel that "if you really want to hurt someone, a weapon is the easiest way to do it" and "when you carry a weapon you get a lot more respect". Note that these questions refer to teenagers in general, rather than to the respondent himself (Appendix B, Table 4). We chose not to ask the more personalized form of the question because the majority of students, and large numbers of dropouts and detainees, were not expected to have ever carried weapons. We wanted all of the respondents to answer these attitude questions. In retrospect, however, we would have preferred to ask this question specifically about guns.

In Toronto, 41% of the students agreed with the statement that using a weapon is the easiest way to hurt someone if you wanted to and over half (52.3%) of their Montreal counterparts felt the same. Twice as many students in Montreal agreed with the assertion that carrying a weapon garnered more respect than in Toronto (31% vs. 15%).

Gun Related Violence: Against self and towards others

Although a very small number of students reported gun related violence to themselves (Toronto N=32; Montreal N=21), a slightly higher proportion of Toronto youth indicated that someone had threatened or tried to hurt them with a gun in their lifetime in comparison to Montreal (7.1% vs. 4.7%) (Appendix B, Table 6). In addition, Toronto students have clearly experienced a greater number of gun related threats in the past year (14.5% reported gun threats to themselves vs. 3.1% in Montreal).

The opposite trend emerges when looking at the commission of gun related violence to others. Again, a small number of students in Toronto (N=13) and Montreal (N=11) reported having ever threatened or tried to hurt someone with a gun in their lifetime. However, among this subset of students, those in Montreal were twice as likely as those in Toronto to indicate threatening or trying to hurt someone with a gun 3 or more times in the past year (22% Toronto vs 51% Montreal).

In Toronto, among those students who reported being the victim of gun related violence (someone threatened or tried to hurt them) 46% also indicated participating in a gang fight in the past year (Appendix B, Table 7). This relationship is even more pronounced in Montreal where 80% of those victimized by gun related violence, also reported engaging in a gang fight in this time frame. Half of the Toronto students who have threatened or tried to hurt someone else with a gun in the past 12 months were also involved in a gang fight during this time and 80% of the students in Montreal report the same. It is important to note that the number of students who have either been threatened with a gun or threatened someone else with a gun is very small. However, it is interesting to note that among this small subset of both samples, a high proportion of them have also been involved in gang fighting.

Accounts of the most violent incident

A separate series of questions asked respondents to identify a specific violent event they had experienced. The degree of seriousness could range from minor (pushing, shoving) to an assault with a weapon. Students who reported having ever experienced a violent event in their lifetime (and some answered that they never had) were asked to answer a series of questions that provided greater detail about the nature and context of the most violent event (if more than one). The following discussion centers on three contextual issues related to the most violent event experienced by the students: gang involvement, gun involvement and drug use and dealing (Appendix B, Table 8).

Gang participation in the event was equally common for students in both cities, although it appeared that gang members were more likely to 'do' the violence than be the target of the incident. Moreover, in spite of the similar reports of gang participation, a greater proportion of students in Montreal attribute the main reason for the incident as being gang related (13% vs 7%).

Guns being used to threaten or try to hurt someone during the incident were reported by twice as many students in Toronto (11%) as in Montreal (6%). However, equal rates of someone actually being hurt with a weapon in this incident (including knife, object or gun) were reported in both cities.

Although not necessarily occurring simultaneously with the violent event, a slightly larger proportion of students in Toronto reported having used any psychoactive substance (including alcohol) on the day of the event (28%) than in Montreal (21%). Finally, less than 10% of the students in both samples indicate the event was in any way related to drug dealing activities.

Vb. Dropouts

Weapon Carrying Behaviour

As seen in Appendix C, Table 9, in comparison to Montreal, twice as many dropouts in Toronto reported 'most' or 'some' of the students in their school carried weapons (50% vs. 25%) and twice as many also indicated having ever carried a weapon to school themselves (46% vs. 22%). An equal number of dropouts in both cities said they knew someone who had brought a gun to school; however, Toronto youth were twice as likely to report having ever carried a gun to school themselves (15% vs. 8%). Although weapons appeared to be more common in the Toronto school environment, similar perceptions of the seriousness of the gun presence in that environment were shared by youth in both cities. Specifically 13% of the Toronto and Montreal dropouts perceived students bringing guns to school as a 'very serious' problem.

A greater proportion of the dropouts in Toronto reported weapon carrying both among their friends and personal weapon carrying. Over three-quarters of the dropouts in Toronto indicated that their friends carried weapons, while 63% reported the same in Montreal, and in Toronto 76% indicated having carried a weapon when they were not in school (in comparison to 38% in Montreal). In both cities, knives emerged as the most frequently reported weapon ever carried (70% in Toronto and 34% in Montreal). Carrying a gun was reported by 33% of the Toronto dropouts and by 18% in Montreal.

The relationships between weapon and gun carrying and gang related activities are presented in Table 10, Appendix C. This table presents a series of analyses to highlight the relationship between the proportion of students who indicated carrying weapons (and specifically guns) who also reported ever participating in a gang fight (and again, these numbers are small).

A stronger relationship emerged between those who reported ever carrying a weapon when they were not at school and gang fighting among the Montreal dropouts. Ever participating in a gang fight was reported by 70% of those who had carried a weapon when they were not at school. This pattern was reported by 43% of those in Toronto. Participation in gang fights also appears to be related to the incidence of carrying a weapon to school for Montreal dropouts. In this case, 90% of those who indicated carrying a weapon to school had also participated in gang fights (46% in Toronto reported the same). Finally, the relationship between having carried a gun to school and gang fighting was evident for dropouts in both Toronto and Montreal. Once again participation in gang fights was greater in Montreal. Among this smaller subset, 88% of the Montreal dropouts also reported participating in a gang fight, as did 71% of the dropouts in Toronto.

Attitudes toward weapons

Respondents in both cities reported carrying a variety of different weapons (Table 9). About two-thirds of the dropouts in Toronto and Montreal agreed that a weapon was the easiest way to hurt someone if you wanted to (66% Toronto and 62% Montreal). Moreover, members of both groups (although slightly greater in Montreal) felt that carrying a weapon gave you more respect (24% Toronto and 30% Montreal).

Gun Related Violence: Against self and towards others

The lifetime experience of gun related violence against self or towards others was greater among Toronto dropouts than those in Montreal (Appendix C, Table 11). Specifically, 44% of the Toronto group reported someone having threatened or tried to hurt them with a gun in their lifetime (vs. 28% in Montreal) and twice as many dropouts in Toronto reported having threatened or tried to hurt someone else with a gun themselves (25% vs. 12%).

Although victimized less often in their lifetime, Montreal dropouts reported a greater amount of gun related victimization in the past 12 months than those in Toronto. One-fifth (21%) of the Montreal dropouts reported being victimized in this manner 3 or more times in the past year compared to 16% in Toronto. Conversely, Toronto dropouts reported being the aggressor of this type of violence more frequently than the dropouts in Montreal. One-quarter of the Toronto group reported threatening or trying to hurt someone with a gun 3 or more times in the past year while 17% of the dropouts in Montreal reported doing so.

Among dropouts who reported being the victim of gun related violence (someone threatened or tried to hurt them) in the past year a slightly higher proportion of those in Montreal also indicate participating in a gang fight in the past year (Appendix C, Table 12). This relationship is even more pronounced in Montreal when examining those dropouts who have threatened or tried to hurt someone else with a gun in the past 12 months. Approximately 88% in Montreal were also involved in a gang fight during this time and 65% of the dropouts in Toronto report the same. The number of dropouts who have threatened someone else with a gun is very small. However, among this small subset of both samples, a high proportion of them have also been involved in gang fighting.

Accounts of the most violent incident

Table 13 reveals that gang involvement was more pronounced in the accounts of the most violent incident the dropouts had experienced in Montreal than in Toronto. A greater proportion of Montreal dropouts reported gang members participating in the violent incident (36% vs. 4%) as well as being the target of the event (18% vs. 1%). Moreover, twice as many youth in Montreal cited 'gang related' as the main cause of the event than they did in Toronto (16% vs. 9%).

Guns being used to threaten or trying to hurt people were reported equally by both groups, however, more dropouts in Montreal indicated that someone was actually hurt with a weapon (knife, gun or object) than in Toronto. Virtually all of the dropouts in both samples said they had used some type of psychoactive substance on the day of the event. Finally, approximately three times as many Montreal dropouts indicated the event was in some way drug related in comparison to those in Toronto (27% vs. 7%).

Gun acquisition

With the evidence presented above that a small minority of dropout youth on the streets of these two large cities are carrying guns, it is of interest to know how they would obtain them. Little information has been available heretofore in Canada pertaining to the acquisition of firearms by youth. Even if these are only perceptions that have not been acted upon, it is still important to gain them.

Tables 14a and 14b in Appendix C provide information on where dropouts say they would most likely go to get a gun if they wanted one, and how long it would take them to get a gun. Among dropouts in both Toronto and Montreal 'buying a gun from a friend or relative' emerged as the most commonly reported gun source (roughly one-third in each city). Following this, 26% of the Toronto dropouts indicated they would not have to pay for a gun but could obtain one from their friends or relatives. One-fifth (20%) of the Toronto dropouts indicated they would get a gun 'on the streets' while 18% of those in Montreal reported 'on the streets' and 'from a drug dealer'. Only 7% of the dropouts in Toronto said they would most likely get one from a drug dealer. Similar proportions of dropouts in both cities indicated they would get a gun legally 'from a store' (though this may reflect ignorance of the limits on under-age purchase).

Overall, Toronto dropouts were more likely to report being able to acquire a gun in a very short period of time. Over half of this group felt they would be able to get a gun in less than one day, in comparison to 38% in Montreal. Moreover, almost twice as many Toronto dropouts indicated they could get a gun in less than one hour (23%) than those in Montreal (13%).

Vc. Detainees

Weapon Carrying Behaviour

Although high rates of weapon carrying are reported by incarcerated male youth in both cities, overall, Toronto detainees reported more weapons carrying than their Montreal counterparts (Appendix D, Table 15). One-quarter (26%) reported most of the students in their school carried a weapon, while 11% of the detainees in Montreal indicated the same. Overall, Toronto detainees perceive a greater degree of weapon carrying in their schools than in the Montreal sample. Although the proportion of detainees in Toronto (77%) who said they knew someone who carried a gun to school was greater than in Montreal, well over half (61%) of the Montreal detainees also said they knew someone who had carried a gun to school. Not surprisingly, Toronto detainees also perceive the problem of students bringing guns to school as being more serious than the detainees in Montreal. Close to one-quarter (24%) of the Toronto detainees said this was a 'very serious' problem compared to 12% in Montreal.

The majority of detainees in both Toronto and Montreal indicated having ever carried a weapon when they were not at school (85% and 76% respectively), and over half of the Toronto detainees reported they have carried a weapon when they were at school (60%). The most commonly reported weapon ever carried by both groups is knives (73% Toronto & 54% Montreal), followed by guns in both cities; however, more Toronto detainees report gun carrying (60% and 49% respectively).

The relationship between weapon and gun carrying and gang involvement was explored in these analyses (Appendix D, Table 16). In both Toronto and Montreal, the majority of those who indicated ever carrying a weapon when they were not at school also reported ever participating in a gang fight (69% in Toronto and 78% in Montreal). A similar pattern emerged when assessing the relationship between weapon carrying to school and gang participation. Among the subset of detainees who carried weapons, 72% in Toronto and 86% in Montreal also said they had engaged in a gang fight. This relationship was particularly pronounced for the small group of detainees who have carried a gun to school. Over 95% of detainees in both cities also reported ever being involved in a gang fight.

Attitudes toward weapons

What proportion of detainees in Toronto and Montreal agree that weapons provide the easiest way to hurt someone and get you more respect? Answers to these questions are provided in Table 15. In Toronto, 70% of the detainees agreed that if you wanted to really hurt someone, a weapon was the easiest way to do it. Half of the group in Montreal also agreed with this statement. Regarding the issue of respect, one-quarter of the detainees in both Toronto and Montreal agreed that if you carried a weapon you were given greater respect by others.

Gun Related Violence: Against self and towards others

Detainees in Montreal and Toronto reported experiencing a considerable amount of gun related violence directed both against themselves and towards others; however, the rates were higher in the Toronto group (Table 17). Almost two-thirds of the detainees in Toronto (61%) indicated someone had threatened or tried to hurt them with a gun in their lifetime and half (49%) of the Montreal group reported the same. Among those detainees who had been threatened with a gun, those in Toronto reported this occurring more frequently than among the Montreal subgroup: two and a half times as many Toronto detainees indicated being threatened with a gun three or more times in the past year than in Montreal (37% vs. 15% respectively).

Similar proportions of detainees in both cities reported threatening or trying to hurt someone else with a gun in their lifetime (46% in Toronto & 41% in Montreal). However, 38% of the Toronto group said they had done this three or more times in the past year in comparison to 24% in Montreal.

Are those involved in some way with gun related violence in the past year also likely to be involved in gang fighting during that same time period? In Toronto, 69% of those threatened by someone with a gun also reported gang fighting during the past year (Table 18). This relationship was more pronounced in Montreal where 80% of those reporting that someone had threatened or tried to hurt them with a gun had also engaged in gang fights over the past 12 months. Those who admitted threatening or trying to hurt someone else with a gun were also generally more likely to report participating in gang fights over the past year as well. Three-quarters of the Toronto detainees who had used a gun to threaten or try to hurt someone in the past year had also engaged in gang fights and 84% of those in Montreal were involved in both these gun and gang activities over the previous year.

Accounts of the most violent incident

Detainees in both cities were asked to describe in detail, the most violent event they had experienced in the past year (Table 19). The following discussion highlights the context of this event as it pertained to gang involvement, gun related violence and drug use and selling.

Overall, gang involvement appeared to be a more prominent characteristic in the most violent event reported by the detainees in Montreal. Within this group 35% said gang members were involved in doing the violence (in comparison to 28% in Toronto) and 23% indicated gang members were the target of this particular event as well (14% in Toronto). Interestingly, although involvement of gangs was reported more frequently in violent incidents in Montreal, an equal proportion of detainees in both sites felt the 'main cause' of the event was gang related (approximately 13%).

Gun involvement in the most serious event was prominent for both Toronto and Montreal detainees. In Montreal, one-third of the detainees reported participants threatened or tried to use a gun to hurt someone and over half (55%) indicated someone was actually hurt with a weapon (knife, object or gun). In Toronto, guns being used to threaten or to try and hurt someone was reported by one-quarter (24%) of the detainees, and just under half (49%) said someone was hurt with a weapon in this incident.

Any psychoactive drug use on the day of the event was reported by virtually all of the detainees in both sites. There appeared to be a stronger degree of drug relatedness to the most violent event reported by the detainees in Montreal than in Toronto. Two-thirds of the Montreal group indicated this event was in some way related to drug dealing activities, while only 13% of the Toronto detainees reported this association.

Gun acquisition

Since two-fifths (40%) of the detainees in both cities reported having threatened or tried to hurt someone with a gun in their lifetime, and about 80% who have ever done so have done so at least once in the past year, how these individuals are currently able to access guns is an important question. When asked where they would most likely get a gun if they wanted one, the top three responses from the Toronto detainees were: buy one from a friend or relative (40%), get one on the streets (20%) and get one from a drug dealer (14%). In Montreal, get one from a drug dealer (28%), buy one from a friend or relative (26%), and get one on the streets (19%) were the top replies. Interestingly, 3% in Toronto and 10% in Montreal said they would get a gun legally from a store (Table 20a).

Finally, in both cities the majority of detainees perceived they could get a gun in less than one day (66% in Toronto and 59% in Montreal) (see Table 20b). Almost two-fifths of the group in Toronto (37%) indicated they could get a gun in less than one hour and 21% in Montreal reported the same.

VI. Multivariate Results - Logistic Regression

The results of a series of hierarchic logistic regression models predicting lifetime gun related violence against self, other, or both, are presented in Appendix E, Tables 21 through 25. The results are reported in odds ratios (ORs) that examine the predictive value of the demographic, custody, school suspension, gang fighting, drug dealing and weapon attitudes variables for these three gun related outcomes. The results are presented separately for the student, dropout and detainee samples and for Toronto and Montreal. While relationships exist between each variable and the outcome of interest, when controlling for the effects of all the variables in the model, only some emerge as having statistically significant effects. The emphasis of these findings will focus on the significant relationships observed. Comments will be also be made to point out other notable non-statistically significant effects, keeping in mind that they could occur by chance.

VIa. Students

Table 21 presents the findings predicting the likelihood of the combined outcome – experience of lifetime gun related violence against either self or others, or both forms - for students in Toronto and Montreal. Having ever participated in a gang fight emerged as the only significant predictor of this outcome in both cities. Students in Toronto who had been involved in a gang fight were 4 times as likely to report experience of gun-related violence than those who had not engaged in gang fighting. This relationship was even more pronounced in Montreal. There, students were almost 11 times more likely to report having being involved in gun related violence when they had taken part in any gang fighting activities.

Positive relationships emerged for the racial indicators in Toronto and Montreal (e.g., Toronto Black OR = 2.08 and Montreal Black OR = 4.30) suggesting that in comparison to Whites, Black youth are more likely to have experienced gun violence. However, when included with the other predictor variables in the model, the effect of race did not significantly influence the likelihood of this outcome. Similarly, while participation in drug selling over the past year is positively associated with gun related violence, as shown in the ORs>1, when considered in combination with the other indicators, drug selling does not have a significant influence.

In addition, non-statistically significant negative relationships are also observed in the student analysis. An odds ratio of .506 in Toronto and .423 in Montreal for the family indicator item suggest that living with both parents is negatively related (reduces the likelihood) of gun involvement for students. But again, this effect does not have a significant overall impact on involvement in gun violence when other factors are controlled.

VIb. Dropouts

Table 22 illustrates the factors that influence the experience of two gun related outcomes – lifetime violence against self and lifetime violence against other - among Toronto dropouts. No significant predictors emerged in column 1 reflecting lifetime gun violence against self, only. Column 2 shows a significant relationship between involvement in the cocaine/crack market and the likelihood of ever being the perpetrator of a violent act involving a gun. The odds of being involved in this type of violent behaviour are 5 times greater for those dropouts who are engaged in selling cocaine and/or crack, in comparison to those not involved in this illicit drug market.

Although not statistically significant, those dropouts in the older age category (16-17 years of age) and those living with both parents were less likely to report both forms of gun related violence. In comparison to 'White' dropouts, a positive, but non statistically significant, relationship did emerge for those who indicated their race as 'Black' and the outcome measure "threatening or trying to hurt someone else with a gun". Therefore, although the race variable may have an impact, being 'Black' per se does not significantly affect the odds of this outcome.

Positive but non-significant relationships were also observed for the items measuring school suspension and lifetime gang fighting among Toronto dropouts. These relationships had the strongest influence on gun violence directed towards others. In addition, although selling cocaine/crack emerged as a significant predictor only for gun violence against another, a positive relationship was found for gun violence against self as well.

Finally, the attitudes items, believing that a weapon is the easiest way to hurt someone and that carrying a weapon gets you more respect, have a very minor though positive, non-significant, relationship with the experience of gun violence against the respondent and towards others.

These analyses were replicated for the dropouts in Montreal and these results are illustrated in Table 23. Reports of any lifetime violence against self is influenced by two factors: (i) being of 'other' racial background and (ii) involvement in the cocaine/crack market. Those of 'other' racial backgrounds were 4.4 times more likely to report lifetime victimization than 'White' dropouts and involvement in the cocaine/crack market dramatically increased the odds of this type of gun related violence. Dropouts who sold cocaine/crack were 11 times as likely to have been the victim of a gun related violent incident in comparison to those not involved in this particular drug market.

No significant predictors emerged in column 2 predicting the likelihood of lifetime gun related violence towards others.

Although not statistically significant, a positive relationship was found between Montreal drop- outs in the older age category and the likelihood of personally experiencing any lifetime gun violence but this effect was reversed for the lifetime violence against another outcome. As with the dropouts in Toronto, while also non-significant, being Black is positively associated with each of the gun related outcomes.

For Montreal dropouts, gang fighting and selling cocaine/crack both revealed positive, non-statistically significant relationships with threatening or trying to hurt someone else with a gun. These effects were quite substantial (OR = 7.00 for gang fighting and OR=4.79 for selling cocaine/crack) but did not have a significant effect on the likelihood of this outcome. Once again, this does not entirely negate the impact these variables have on gun related violence but rather suggests that there are other predictors (not identified in this model) that may play a more significant role influencing this outcome.

VIc. Detainees

Tables 24 and 25 present the findings predicting lifetime gun related violence outcomes among Toronto and Montreal detainees. Some similar predictors emerge for both outcomes among Toronto detainees (Table 24). Column 1 shows a significant relationship between ever participating in a gang fight and having ever sold drugs with lifetime gun related violence against self. Detainees involved in a gang fight were 3.2 times as likely to report this type of gun related violence than those who had not been involved in a gang fight. Moreover, the odds of this involvement were 3.7 times greater for those having any drug market participation than that of those not involved in drug selling activities.

In addition, agreeing with the perception that the use of a weapon is the easiest way to hurt someone if you wanted to, increased the likelihood of lifetime gun related violence against self. Those detainees who agreed with this statement about weapons were more than two and a half times (OR = 2.6) more likely to report gun violence against themselves than those who did not ascribe to this belief.

Finally, column 2 reveals three significant predictors of lifetime gun related violence towards others. As with the other outcome, gang fighting participation (OR = 3.2) and drug selling (OR = 5.3) significantly influence the likelihood of this outcome. Those involved in gang fighting are three times more likely, and those selling drugs are more than 5 times as likely, to report threatening or trying to hurt someone with a weapon than those detainees not engaged in these activities. In addition, this model reveals a significant influence of dealing in the cocaine/crack market on the likelihood of gun aggression towards others. Specifically, detainees involved in selling these particular drugs are almost 4 times more likely to have threatened or tried to hurt someone with a gun.

Although none of the background demographic factors had a statistically significant effect on the gun related violent outcomes, the direction of the relationships is worth comment. A positive relationship was observed between age and gun violence against oneself. However, this relationship was negative for gun violence against another. Thus, while age per se does not significantly affect these outcomes, there is something about this variable that affects the gun violence outcomes differently. Self-identified racial backgrounds of 'Black' and 'Asian' also emerged as having a positive, non-significant effect on threatening or trying to hurt someone with a gun. Again, while these demographic indicators do not have a significant effect on this outcome, this influence may contribute to gun related violence through other indicators in the model. Lastly, for each gun related violence outcome, living with both parents had a negative, non statistically significant, effect.

Another point is that positive relationships were observed for the two attitudinal items and lifetime gun violence against another. Although these were not significant effects, these findings do suggest that attitudes may be related to violent behaviour. This appeared to be particularly the case for the belief that carrying a weapon got you more respect (OR = 3.3 non-significant). Understanding what it is about this belief or other pro weapons attitudes that may be related to actual behaviour could be probed in future studies.

The final table, 25, illustrates the significant predictors among the detainees in Montreal. Participation in gang fights emerged as an important predictor of gun related violence outcomes. In comparison to those not involved in gang fighting, Montreal detainees who had been involved in a gang fight were 3 times more likely to have been the victim only of this type of event and 3 and half times as likely to indicate having threatened or tried to hurt someone else with a gun in their lifetime.

The only other significant predictor to emerge in this series of analyses is shown in column 1 predicting lifetime gun violence against self. In this model believing that carrying a weapon gets you more respect significantly reduced the odds of being a victim of a gun related incident. It is possible this respect item taps into masculinity and sensitivity to expressions of disrespect; do those who do carry a weapon to get respect intimidate those who might be tempted to attack them, and thus escape victimization? Or, is it possible that this perception is more closely aligned with those who do not adopt the guns/gangs way of life, and avoid being a victim through a strategy of avoidance? At this point, we can only speculate.

As with the detainees in Toronto, none of the demographic indicators emerged as significant predictors of the gun related outcomes for Montreal detainees; however, a few comments about the observed relationships will be noted. In comparison to being White, Black detainees showed a negative relationship with reports of lifetime gun related violence perpetrated against themselves. However, this relationship was reversed when considering gun related violence against another. Once again while race per se may not have a significant effect on these violent outcomes, it may influence behaviour through other indicators. Thus, while race should not be ignored, more attention needs to be given to what other factors associated with being a member of a particular racial or ethnic group affect gun related violent outcomes.

VII. Executive Highlights

VIIa. Descriptive Analysis - Toronto-Montreal Comparisons

Three Samples – Male students, dropouts, detainees – General impressions

Students

Dropouts

Detainees

VIIb. Multivariate Logistic Regression

Students

Dropouts

Detainees

A Note on Girls

VIII. Recommendations

As expected, the highest rates of weapons involvement and gun violence are found in detainees, followed by dropouts, and the lowest rates are among students. This suggests that different forms of intervention should be targeted to the different groups. For students the most high-risk members appear to be those with histories of gang fighting, with no other factors of comparably strong impact. Thus, preventive efforts aimed at providing alternatives to youth before they seek gang liaisons should be a high priority. It is also evident that every attempt must be made to keep youth in school and to provide productive opportunities for them to succeed.

For those who have dropped out, it is vital to attract them back to classes or provide other alternative schooling. The data suggest that dropout youth who get involved in selling cocaine/crack are at higher risk of being involved in gun related violence of various sorts. This suggests that providing economic, not just recreational, options, are important for youth who may be at the brink of greater criminal involvements.

For the detainees, who have already committed serious enough crimes (though we do not record them) to receive custody dispositions, this may be the last chance for interventions to prevent full scale adult criminal careers. For only the detainees were the attitudinal items towards weapons significant, suggesting that they may have developed ways to 'neutralize' their more violent behaviors, compared to the other two less delinquent groups of youth. Such techniques to justify violent or predatory behaviour that is at odds with conventional societal values has been documented in several studies as a common feature of delinquent youth.7 Little is known specifically about how this process might be related to firearms.

Along with the importance of gang fighting histories in both cities and drug selling (especially cocaine/crack in Toronto), as predictors of gun violence in the detained youth, the picture presented is of youth whose escalation into more serious crime is accompanied by more reliance on weapons. This interpretation is supported by a US longitudinal study that indicated youth recourse to guns intensified at a later stage of the deviant trajectory, after early delinquency and gang involvements, and accompanying progression to drug dealing8. While every effort should be made to re-orient youth in custody to other goals, through education and counseling about the long term risks of a criminal path, it is also reasonable to see that enforcement must play an important role in identifying such violence-prone youth and removing them from the street.

There are many variables that have not been shown to have a significant effect on the gun violence outcomes, despite popular views and expectations. For example, drug use itself, and lack of an intact family, as well as racial background, were generally not important once all the variables of interest were controlled. Race should not be ignored but more attention needs to be given to what other factors associated with being a member of a particular racial or ethnic group affect gun related violent outcomes, as they appear also to vary by group and city.

As well, though our measure of gang fighting was a very powerful predictor in many of the models, the variation between Toronto and Montreal requires further examination. The descriptive data show that Montreal has an overall higher gang fighting prevalence while Toronto displays generally higher gun exposure. Indeed most descriptive indicators converge in showing more gun carrying, threat, use and concern about guns, among Toronto youth in all three samples. This is particularly important among students, which as the most representative sample of youth in general, provides the best reflection of the potential for gun violence in the adolescent population. Also detainees and dropouts in Toronto indicated they could acquire a gun more quickly, on the average, than those in Montreal. Since the multivariate analysis showed a stronger association of guns with gangs in Montreal, and a stronger association of guns with drug dealing in Toronto, it would follow that the "gun culture" of cities varies for youth, and appropriate interventions must be considered separately.

Clearly, more research is needed about the context of gun related violent events. The sequence, and the distinction between victim and perpetrator, may arise in a variety of spontaneous occurrences. The distinction between defensive carrying, status enhancement, and plans for aggressive use (e.g. in crimes) is also an important topic for further examination.

Funding for other studies that tap into why youth feel the 'need' to arm themselves, its relationship to gang involvement (and how youth in general see gang involvement), involvement in drug selling, and detailed questions about gun acquisition, would all provide valuable additional data to guide intervention efforts.

Appendix A: Background Characteristics

Table 1: Background Characteristics – Toronto & Montreal Students
Characteristic Boys Only Toronto (N=456)
%		 Montreal (N=448)
%
Age 14 or younger 16.3 30.5
15 18.9 27.0
16 27.9 23.5
17 or older 36.9 18.9
Race White 45.3 74.6
Black 9.9 9.1
Asian 26.1 5.0
Native Canadian 0.0 0.1
Hispanic 3.9 3.5
Mixed 12.7 7.1
Other 2.0 0.6
Lives with both parents Yes 79.3 76.0
School Suspension Yes 8.0 26.0
Gang Fighting Ever taken part in gang fight – yes 16.2 (n=71) 23.6 (n=100)
Gang fight before 13 – yes 10.0 19.1
Take part in gang fight in last 12 months 11.9 (n=52) 19.9 (n=84)
How many times?
0 88.1 80.0
1 4.0 12.4
2 3.3 1.2
3-9 3.9 4.6
10+ 1.5 1.9
Drug Selling Sold drugs past 12 months 15.3 10.3
Contact with justice system A night in jail/detention 2.8 5.9
Table 2: Background Characteristics – Toronto & Montreal Dropouts
Characteristic Boys Only Toronto (N=116)
%		 Montreal (N=102)
%
Age 14 or younger 8.6 8.8
15 13.8 13.7
16 30.2 15.7
17 or older 47.4 61.8
Race White 43.1 60.8
Black 28.4 17.6
Asian 6.0 0.3
Other 22.4 21.6
Lives with both parents Yes 12.1 29.4
School Suspension Yes 91.4 83.2
Gang Fighting Ever taken part in gang fight – yes 37.9 (n=44) 59.8 (n=61)
Gang fight before 13 – yes 22.4 (n=26) 38.2 (n=39)
How many times in the past 12 months have you taken part in a gang fight?
0 69.8 49.0
1 13.8 14.7
2 6.9 6.9
3-9 6.9 15.7
10+ 2.6 13.7
Drug Selling Ever sold drugs 69.0 51.0
Sell cocaine/crack or both 20.7 9.8
Contact with justice system Have you ever been locked up? 71.9 20.6
A night in jail/detention 69.2 189.5
Table 3: Background Characteristics – Toronto & Montreal Male Detainees
Characteristic Boys Only Toronto (N=132)
%		 Montreal (N=146)
%
Age 14 or younger 10.6 0.7
15 27.3 8.2
16 31.1 26.7
17 or older 31.1 64.4
Race White 54.5 71.2
Black 26.5 13.
Asian 4.5 0.0
Other 14.4 15.1
Lives with both parents Yes 24.2 20.
School Suspension Yes 99.2 95.9
Gang Fighting Ever taken part in gang fight – yes 64.6 (n=82) 73.0 (n=103)
Gang fight before 13 – yes 39.4 42.1
How many times in the past 12 months have you taken part in a gang fight?
0 43.2 36.6
1 12.9 12.4
2 8.3 11.7
3-9 18.2 20.0
10+ 17.4 19.3
Drug Selling Ever sold drugs 84.0 82.8
Sell cocaine/crack or both 40.2 41.8
Contact with justice system Average number of days in custody 117 502

Appendix B: Descriptive Findings – Students

Table 4: Weapon Carrying: Toronto & Montreal Students
  Toronto (N=456)
%		 Montreal (N=448)
%
In your school, how many students carry some sort of weapon?
Most 4.5 1.9
Some 27.3 10.5
A few 52.2 45.5
None 16.0 42.2
Do any of your friends carry a weapon?
Yes 28.0 23.6
If you carry a weapon, what weapon(s) do you carry?
Mace 4.0 3.2
Knife 20.4 17.5
Gun 4.2 2.8
Have you ever carried a weapon when you were not in school?
Yes 40.0 41.2
Have you ever carried a weapon when you were in school?
Yes 15.1 18.7
In your school, how serious a problem is students bringing guns to school?
Very serious 21.1 12.8
Somewhat serious 12.3 6.0
Not at all serious 66.6 81.1
Do you know anyone who has carried a gun to school?
Yes 21.5 7.2
Have you ever carried a gun to school?
Yes 0.8 2.2
If you really want to hurt someone a weapon is the easiest way to do it
Agree 41.0 52.3
When you carry a weapon you get a lot more respect
Agree 14.6 30.7
Table 5. Proportion of male students who report carrying a weapon/gun carrying in and out of the school environment who also report ever taking part in a gang fight.
Toronto Montreal
  % yes   % yes
Ever carry a weapon when not at school (N=171) Ever carry a weapon when not at school (N=173)
Ever taking part in a gang fight 36.8% Ever taking part in a gang fight 43.9%
Ever carry a weapon when at school (N=67) Ever carry a weapon when at school (N=77)
Ever taking part in a gang fight 43.3% Ever taking part in a gang fight 49.4%
Ever carry a gun (N=18) Ever carry a gun (N=12)
Ever taking part in a gang fight 61.1% Ever taking part in a gang fight 75.0%
Ever carry a gun to school (N=4) Ever carry a gun to school (N=9)
Ever taking part in a gang fight 25.0% Ever taking part in a gang fight 44.4%
Table 6: Experience of gun related violence: Against oneself and towards others
  Toronto (N=456)
%		 Montreal (N=448)
%
Has someone ever threatened or tried to hurt you with a gun?
Yes 7.1 (n=32) 4.7 (n=21)
Among those who say yes, how many times has someone threatened or tried to hurt you with a gun in the past 12 months?
0 13.0 5.1
1 70.4 81.6
2 2.0 10.2
3+ 14.5 3.1
Have you ever threatened or tried to hurt someone with a gun?
Yes 3.0 (n=13) 2.5 (n=11)
Among those who say yes, how many times have you threatened or tried to hurt someone with a gun in the past 12 months?
0 2.0 17.3
1 50.5 18.8
2 25.8 13.3
3+ 21.7 50.6
Table 7: Proportion of male students who report past 12 month gun related violence against themselves and gun related violence against another who also report taking part in a gang fight in the past year
Toronto Montreal
12 month gun related violence against oneself (N=22) 12 month gun related violence against oneself (N=10)
Gang fight in past 12 months 45.5% Gang fight in past 12 months 80.0%
Past 12 month gun violence against others
(N=8)		 Past 12 month gun violence against others
(N=80)
Gang fight in past 12 months 50% Gang fight in past 12 months 50%
Table 8: Accounts of the Most Violent Incident
  Toronto (N=285)
%		 Montreal (N=381)
%
Did gang members participate in the violence?
Yes 21.7 20.1
Were gang members the target of the violence?
Yes 9.2 8.7
Were guns involved in this violent incident (i.e. threatening or trying to hurt)
Yes 10.7 5.8
Did someone get hurt with a knife, object or gun during this incident?
Yes 9.6 9.7
Had you used any psychoactive substances on the day of this event?
Yes 27.2 22.8
Was this event in any way related to drug dealing activities?
Yes 6.0 7.3
Was the main cause of this event gang related?
Yes 6.6 12.9

Appendix C: Descriptive Findings – Dropouts

Table 9: Weapon Carrying: Toronto & Montreal Male Dropouts
  Toronto (N=116)
%		 Montreal (N=102)
%
In your school, how many students carry some sort of weapon?
Most 19.8 7.0
Some 30.2 18.0
A few 43.1 45.0
None 6.9 30.0
Do any of your friends carry a weapon?
Yes 78.4 62.7
If you carry a weapon, what weapon(s) do you carry?
Mace 18.4 7.8
Martial arts weapon 15.8 4.9
Club 28.1 16.7
Knife 70.2 34.3
Gun 32.5 17.6
Have you ever carried a weapon when you were not in school?
Yes 75.9 38.2
Have you ever carried a weapon when you were in school?
Yes 45.7 21.6
In your school, how serious a problem is students bringing guns to school?
Very serious 13.8 12.7
Somewhat serious 25.0 20.6
Not at all serious 61.2 66.7
Do you know anyone who has carried a gun to school?
Yes 56.9 57.8
Have you ever carried a gun to school?
Yes 14.7 7.8
If you really want to hurt someone a weapon is the easiest way to do it
Agree 65.5 61.8
When you carry a weapon you get a lot more respect
Agree 24.1 30.4
Table 10. Proportion of male dropouts who report carrying a weapon/gun carrying in and out of the school environment who also report ever taking part in a gang fight.
Toronto Montreal
  % yes   % yes
Ever carry a weapon when not at school (N=191) Ever carry a weapon when not at school (N=64)
Ever taking part in a gang fight 42.9% Ever taking part in a gang fight 70.3%
Ever carry a weapon when at school (N=88) Ever carry a weapon when at school (N=39)
Ever taking part in a gang fight 45.5% Ever taking part in a gang fight 89.7%
Ever carry a gun (N=37) Ever carry a gun (N=18)
Ever taking part in a gang fight 62.2% Ever taking part in a gang fight 88.9%
Ever carry a gun to school (N=17) Ever carry a gun to school (N=8)
Ever taking part in a gang fight 70.6% Ever taking part in a gang fight 87.5%
Table 11: Experience of gun related violence: Against oneself and towards others
  Toronto (N=116)
%		 Montreal (N=102)
%
Has someone ever threatened or tried to hurt you with a gun?
Yes 44.0 (n=51) 28.4 (n=29)
Among those who say yes, how many times has someone threatened or tried to hurt you with a gun in the past 12 months?
0 33.3 10.3
1 41.2 58.6
2 9.8 10.3
3+ 15.7 20.7
Have you ever threatened or tried to hurt someone with a gun?
Yes 25.0 (n=29) 11.8 (n=12)
Among those who say yes, how many times have you threatened or tried to hurt someone with a gun in the past 12 months?
0 39.3 33.3
1 21.4 33.3
2 14.3 16.7
3+ 25.0 16.7
Table 12: Proportion of male dropouts who report past 12 month gun related violence against themselves and gun related violence against another who also report taking part in a gang fight in the past year
Toronto Montreal
12 month gun related violence against oneself (N=34) 12 month gun related violence against oneself (N=26)
Gang fight in past 12 months 52.9% Gang fight in past 12 months 57.7%
Past 12 month gun violence against others
(N=17)		 Past 12 month gun violence against others
(N=8)
Gang fight in past 12 months 64.7% Gang fight in past 12 months 87.5%
Table 13: Accounts of the Most Violent Incident
  Toronto (N=115)
%		 Montreal (N=99)
%
Did gang members participate in the violence?
Yes 4.4 36.1
Were gang members the target of the violence?
Yes 0.9 18.1
Were guns involved in this violent incident (i.e. threatening or trying to hurt)
Yes 14.0 12.0
Did someone get hurt with a knife, object or gun during this incident?
Yes 17.5 26.5
Had you used any psychoactive substances on the day of this event?
Yes 100.0 100.0
Was this event in any way related to drug dealing activities?
Yes 7.0 26.5
Was the main cause of this event gang related?
Yes 8.8 15.7
Table 14a: Obtaining guns: Where would you most likely get a gun if you wanted one?
  Toronto (N=116)
%		 Montreal (N=102)
%
Steal one 3.4 3.9
Trade something for one 1.7 7.8
Buy one from a friend or relative 31.9 34.3
Get one from a drug dealer 6.9 17.6
Get one on the streets 19.8 17.6
Get one from friend or relative (don't have to pay)* 26.7 0.0
Get one legally – from a store* 4.3 5.9
Other 5.2 12.7

* these response categories were extracted from the 'other' response option to provide greater details

Table 14b: Obtaining guns: How long would it take you to get a gun if you wanted one?
  Toronto (N=116)
%		 Montreal (N=102)
%
<1 hour 23.3 12.7
1-3 hours 12.1 9.8
4 – 24 hours 19.8 15.7
1-2 days 16.4 18.6
More than 2 days 18.1 10.8
I don't know 10.3 32.4

Appendix D: Descriptive Findings - Detainees

Table 15: Weapon Carrying: Toronto & Montreal Male Detainees
  Toronto (N=456)
%		 Montreal (N=448)
%
In your school, how many students carry some sort of weapon?
Most 25.8 10.8
Some 39.4 21.6
A few 28.0 46.0
None 6.8 21.6
Do any of your friends carry a weapon?
Yes 90.9 84.7
If you carry a weapon, what weapon(s) do you carry?
Mace 9.9 10.4
Martial Arts weapon 12.2 3.5
Club 28.2 25.0
Knife 72.5 54.2
Gun 60.3 48.6
Have you ever carried a weapon when you were not in school?
Yes 84.7 75.9
Have you ever carried a weapon when you were in school?
Yes 59.5 42.1
In your school, how serious a problem is students bringing guns to school?
Very serious 23.5 11.6
Somewhat serious 37.1 22.6
Not at all serious 39.4 65.8
Do you know anyone who has carried a gun to school?
Yes 77.3 61.4
Have you ever carried a gun to school?
Yes 30.6 18.6
If you really want to hurt someone a weapon is the easiest way to do it
Agree 69.7 51.0
When you carry a weapon you get a lot more respect
Agree 24.2 24.8
Table 16. Proportion of male detainees who report carrying a weapon/gun carrying in and out of the school environment who also report ever taking part in a gang fight.
Toronto Montreal
  % yes   % yes
Ever carry a weapon when not at school (N=115) Ever carry a weapon when not at school (N=118)
Ever taking part in a gang fight 68.7% Ever taking part in a gang fight 78.0%
Ever carry a weapon when at school (N=106) Ever carry a weapon when at school (N=106)
Ever taking part in a gang fight 71.7% Ever taking part in a gang fight 85.8%
Ever carry a gun (N=74) Ever carry a gun (N=67)
Ever taking part in a gang fight 79.7% Ever taking part in a gang fight 85.1%
Ever carry a gun to school (N=36) Ever carry a gun to school (N=26)
Ever taking part in a gang fight 92.2% Ever taking part in a gang fight 96.2%
Table 17: Experience of gun related violence: Against oneself and towards others
  Toronto (N=132)
%		 Montreal (N=146)
%
Has someone ever threatened or tried to hurt you with a gun?
Yes 61.4 (n=81) 49.3 (n=72)
Among those who say yes, how many times has someone threatened or tried to hurt you with a gun in the past 12 months?
0 28.4 31.9
1 29.6 29.0
2 4.9 24.6
3+ 37.0 14.5
Have you ever threatened or tried to hurt someone with a gun?
Yes 46.2 (n=61) 41.1 (n=60)
Among those who say yes, how many times have you threatened or tried to hurt someone with a gun in the past 12 months?
0 21.3 19.0
1 24.6 41.4
2 16.4 15.5
3+ 37.7 24.1
Table 18: Proportion of male detainees who report past 12 month gun related violence against themselves and gun related violence against another who also report taking part in a gang fight in the past year
Toronto Montreal
  % yes   % yes
12 month gun related violence against oneself (N=54) 12 month gun related violence against oneself
(N=45)
Gang fight in past 12 months 68.5 Gang fight in past 12 months 80.0
Past 12 month gun violence against others (N=43) Past 12 month gun violence against others (N=44)
Gang fight in past 12 months 76.7 Gang fight in past 12 months 84.1
Table 19: Accounts of the Most Violent Incident
  Toronto (N=132)
%		 Montreal (N=142)
%
Did gang members participate in the violence?
Yes 27.5 35.2
Were gang members the target of the violence?
Yes 13.7 22.8
Were guns involved in this violent incident (i.e. threatening or trying to hurt)
Yes 24.4 29.2
Did someone get hurt with a knife, object or gun during this incident?
Yes 48.9 55.4
Had you used any psychoactive substances on the day of this event?
Yes 100.0 100.0
Was this event in any way related to drug dealing activities?
Yes 13.0 40.9
Was the main cause of this event gang related?
Yes 13.7 12.9
Table 20a: Obtaining guns: Where would you most likely get a gun if you wanted one?
  Toronto (N=132)
%		 Montreal (N=146)
%
Steal one 9.1 2.8
Trade something for one 1.5 1.4
Buy one from a friend or relative 40.2 26.2
Get one from a drug dealer 13.6 27.6
Get one on the streets 19.7 18.6
Other 15.9 23.4
Table 20b: Obtaining guns: How long would it take you to get a gun if you wanted one?
  Toronto (N=132)
%		 Montreal (N=146)
%
<1 hour 37.1 20.7
1-3 hours 12.9 17.9
4 – 24 hours 15.9 20.0
1-2 days 18.9 15.2
More than 2 days 12.9 18.6
I don't know 2.3 7.6

Appendix E: Multivariate Logistic Regression Analysis

Table 21: Logistic regression predicting the likelihood of lifetime gun related violence against self and/or towards others among students in Toronto and Montreal
  Toronto
Lifetime gun violence against self, other or both
OR		 Montreal
Lifetime gun violence against self, other or both
OR
16 - 17 years old .889 1.07
Black 2.08 4.30
Asian 2.33 0.00
Other 1.92 2.84
Live with both parents .506 .423
Been in custody (ever) .838 1.92
Been suspended (ever) .910 1.55
Participated in a gang fight (ever) 4.15** 10.87**
Sold drugs in the past 12 months 1.89 1.39
Believe a weapon is the easiest way to hurt someone 1.19 1.09
Believe carrying a weapon gets you more respect 1.69 .773
Table 22: Logistic regression predicting the likelihood of gun related violence outcomes among Toronto Dropouts
  Lifetime against self
OR		 Lifetime against other
OR
16 - 17 years old .324 .713
Black .767 1.33
Asian .000 .000
Other .513 .673
Live with both parents .378 .280
Been suspended (ever) 1.73 7.68
Participated in a gang fight (ever) 2.34 2.68
Sold drugs (ever) .965 4.25
Sold cocaine/crack 3.18 5.05*
Believe a weapon is the easiest way to hurt someone 2.44 1.01
Believe carrying a weapon gets you more respect 1.05 1.30
Table 23: Logistic regression predicting the likelihood of gun related violence outcomes among Montreal Dropouts
  Lifetime against self
OR		 Lifetime against other
OR
16 - 17 years old 1.83 .843
Black 1.65 1.44
Asian na na
Other 4.36* 4.32
Live with both parents 1.17 1.09
Been suspended (ever) .510 1.06
Participated in a gang fight (ever) .829 7.00
Sold drugs (ever) 1.21 .986
Sold cocaine/crack market 11.15* 4.79
Believe a weapon is the easiest way to hurt someone 2.22 .587
Believe carrying a weapon gets you more respect .369 .735
Table 24: Logistic regression predicting the likelihood of gun related violence outcomes among Toronto Detainees
  Lifetime against self
OR		 Lifetime against other
OR
16 - 17 years old 1.95 .705
Black .535 1.71
Asian .476 2.64
Other .466 .189*
Live with both parents .536 .294
Been suspended (ever) .000 .000
Participated in a gang fight (ever) 3.16* 3.22*
Sold drugs (ever) 3.70* 5.29*
Involved in cocaine/crack market 2.81 3.77**
Believe a weapon is the easiest way to hurt someone 2.63* 1.67
Believe carrying a weapon gets you more respect 1.27 3.28
Table 25: Logistic regression predicting the likelihood of gun related violence outcomes among Montreal Detainees
  Lifetime against self
OR		 Lifetime against other
OR
16 - 17 years old 1.86 .682
Black .444 2.22
Asian na na
Other 1.37 1.31
Live with both parents .806 .692
Been suspended (ever) 1.74 2.66
Participated in a gang fight (ever) 2.99* 3.51*
Sold drugs (ever) 1.43 1.16
Involved in cocaine/crack market 1.56 2.05
Believe a weapon is the easiest way to hurt someone 1.16 .594
Believe carrying a weapon gets you more respect .404* .988

Footnotes

1 Erickson P., Butters, J., et al. (in press) Girls and Weapons: An International Study of the Perpetration of Violence. Journal of Urban Health.2006.

2 Drugs, Alcohol and Violence International [DAVI] was funded by the National Institute of Drug Abuse [NIDA] through grant # RO1-DA11691-01A1

3 Funded by the Centre National de Prevention du Crime grant #3150-U4 and the Social Sciences and Humanities Research Council of Canada grant #410-2002-1154

4 Paglia A, Adlaf E M. Secular trends in self-reported violent activity among Ontario students, 1983-2001. Can J Public Health. 2003;94: 212-217.; Sheley J, Wright J. In the Line of Fire: Youth, Guns and Violence in Urban America. New York: Aldine de Gruyer; 1995.

5 Chesney-Lind M.Girls and Violence: Is the gender gap closing? National Electronic Network on Violence Against Women. 2004:1-8. http://www.vawnet.org/DomesticViolence/Research/VAWnetDocs/AR_GirlsViolence.pdf

6 Tanner, J. & Wortley, S. (2002). The Toronto Youth Crime & Victimization Survey. Toronto: Centre of Criminology.

7 Copes, H. Societal attachments, offending frequency, and techniques of neutralization. Deviant Behavior. 2003; 24:535-550

8 Lizotte, AJ, Krohn, MD, Howell, JC, Tobin, K, Howard, GJ. Factors influencing gun carrying among young urban males over the adolescent-young adult life course. Criminology. 2000; 38:811-834

Date modified: