Show Summary Details

Page of

PRINTED FROM OXFORD HANDBOOKS ONLINE ( © Oxford University Press, 2018. All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a title in Oxford Handbooks Online for personal use (for details see Privacy Policy and Legal Notice).

Subscriber: null; date: 20 February 2019

Case Study: Black Homicide Victimization in Toronto, Ontario, Canada

Abstract and Keywords

Most criminological theory and research on the black homicide victimization is grounded in the American context, which raises important generalizability issues given the exceptional level of lethal violence that is used as the standard in this inquiry. This case study examines the social and spatial distribution of black homicide victimization in Toronto, Ontario, Canada between 1988 and 2003. Results suggest that, as in American cities, blacks in Toronto are over-represented as homicide victims and offenders, but there are important differences in the spatial distribution and ecological correlates of this violence. These findings highlight the importance of cross-national research when investigating the generalizability of findings from U.S.-based research on racially disaggregated homicide rates.

Keywords: homicide, race, neighborhoods and crime, spatial analysis, comparative criminology

Almost three decades ago, Archer and Gartner (1984) wrote that much criminological knowledge on the nature and causes of violent crime, including homicide, is “disconcertingly provincial,” due in part to the relative dominance of research and theory grounded in the American context. One consequence on the intra-national focus of this inquiry is that the degree to which theory and research on the causes and correlates of violent crime “holds” across international boundaries cannot be ascertained (Farrington 2000; LaFree 2007; Stamatel 2009). The issue of generalizability looms especially large in the literature on the social and spatial distribution of urban homicide, which rests almost exclusively on American data. That a preponderance of this inquiry has been conducted within the boundaries of a single society is especially concerning, given the exceptional level of lethal violence that is used as the standard in this inquiry and upon which criminological theory and research is premised. The dearth of Canadian theory and research on urban homicide is understandable, however, in light of a set of institutional and political barriers that make it extremely difficult for Canadian criminologists to gain access to the requisite data.

From the American experience, we know that the social distribution of homicide—the distribution of risk across social groups—is heavily influenced by race,1 with African Americans, especially young African-American males particularly vulnerable to high levels of homicide victimization and offending. For example, although African Americans accounted for 13 percent of the U.S. population in 2005, they were victims in approximately half of the nation’s homicides. Furthermore, when African-American homicide rates are disaggregated by the sex and age of victims, we see extraordinary concentrations in the distribution of risk among young males (15–29) and that these killings were almost exclusively committed by other African Americans (93 percent) (Harrell 2007). We also know that clustering in the social distribution of homicide by race is mirrored spatially, such that neighborhoods that are (p. 431) home to large numbers of poor, African-American residents tend also to experience high levels of lethal violence (Sampson 1987; Kubrin and Wadsworth 2003; Peterson and Krivo 2005; Xie 2010).

Little is known about the relationship between race and homicide in the Canadian context. Canadian justice agencies are loath to acknowledge and discuss the intersection of race, crime, and the criminal justice system—and this is perhaps most especially the case when it comes to violent crime. This “erasure of race” (Jiwani 2002) from criminal justice discourse and a concomitant ban on the release of racially disaggregated criminal justice data makes it very difficult for researchers in Canada to document and examine issues related to racial disparity in the risk of lethal violence. At the same time, however, the concept of race is commonly invoked in public and media discourse in Canada, perhaps most especially with respect to issues related to crime and violence (Mosher 1998; Henry and Tator 2000; Ezeonu 2008).

As a consequence, I find myself, a Canadian criminologist studying and teaching at an inner-city university in Toronto, exporting theory and research grounded in the American context to help understand what appears to be a similar social and spatial clustering of the risk of homicide among poor black Torontonians and the urban neighborhoods in which they live.2 Doing so necessarily requires that I reiterate the important caveat that criminological research and theory on the African-American experience may—or may not—be helpful in understanding the experience of being poor and black in Canada and the ways in which this may shape the risk of lethal violence in Canadian cities.3 We simply don’t know—there are significant gaps in our knowledge base in this regard. It is likely, however, unwise to assume that there is a core urban “black” experience that transcends national borders, even if they are as economically and socially porous as that which separates Canada and the United States. An important first step toward building a more cross-national understanding of the connections between race and urban violence, then, is to document the existence and extent of racial disparity in homicide victimization outside of the American context. That is the purpose of this case study, which will examine the social and spatial distribution of black homicide in Toronto, Canada’s largest city, for the period 1988–2003.

The main conclusions of this case study include:

  • Although the overrepresentation of blacks as homicide victims and offenders has received much empirical attention in the United States, the issue has received comparatively little academic and public policy consideration in Canada.

  • The dearth of research on black homicide in Canada stems, in large part, from barriers to gaining access to official homicide data, in general, and racially disaggregated homicide data in particular. This makes it extremely difficult for researchers to document and examine, and for policy makers to address, the issue.

  • As a consequence of the suppression of official criminal justice data on homicide, the news media plays a crucial role in constructing the issue of black homicide victimization and in shaping popular understandings of this violence in Canada.

  • (p. 432) Research suggests that public policy responses are also shaped by media constructions of the causes and correlates of urban violence. This is concerning in light of the fact that news media coverage has been shown to negatively stereotype, criminalize, and misrepresent the black community. The development of effective violence prevention and control policies requires thorough and systematic empirical examination of the nature of the problem at which these policies are aimed.

  • Cross-national research that examines the disproportionate risk of homicide experienced by disadvantaged minority groups holds great potential to expand the breadth and depth of criminological theory and research on urban violence and to contribute to a growing knowledge base of effective public policy aimed at violence prevention and control.

This essay begins with a discussion of what we know about black homicide in Canada, which, to date, is largely informed by the news media. The second section outlines the ways in which researchers in Canada may request homicide data and my experience with this process in Toronto. Given the de facto ban on the release of race-based criminal justice data in Canada, section III discusses the process by which I gathered information on the racial/ethnic background of victims and offenders in my sample. In section IV, I present an overview of my central findings, which highlight the perhaps ironic conclusion that despite the official silence on issues of race and violent crime in Canada, “race matters” for understanding the social and spatial distribution of black homicide victimization. The final section discusses directions for future research and major policy implications that stem from my findings.

I. What We Know about Black Homicide Victimization in Canada

Although race-based data on victims and offenders of violent crime, including homicide, are routinely collected and used internally by agents of the criminal justice system, in Canada these data are not made publicly available.4 There is, however, one exception: Aboriginal identity data is collected and disseminated (albeit not in an entirely systematic way) by many criminal justice agencies in Canada, due to what amounts to limited permission in a political culture that is otherwise not open to discussions of race and crime.5 This permission appears to stem from a collective understanding among researchers, policy makers, and others that a series of historic, social, and economic conditions have, taken together, played a key role in the overrepresentation of Aboriginal Canadians within the criminal justice system, as both victims and offenders (Brzozowski, Taylor-Butts, and Johnson 2006).6 As such, the vast majority of empirical evidence on the issue of racial disparity in homicide victimization is specific to Canada’s Aboriginal population.

(p. 433) Canadian research shows that although Aboriginal people constituted, on average, about 3 percent of the Canadian population between 1997 and 2004, they represented 17 percent of homicide victims in Canada (Brzozowski, Taylor-Butts, and Johnson 2006). We also know that the distribution of risk varies according to victim sex: between 1997 and 2000, the average victim homicide rate for Aboriginal males was 12.2 per 100,000 population, which was twice that of Aboriginal females (5.4 per 100,000), and almost seven times greater than the rate for non-Aboriginal male victims (1.8 per 100,000). The context for and characteristics of Aboriginal homicide victimization also differed from that of non-Aboriginals: Aboriginal homicides were more likely to involve stabbings (44 percent) or beatings (32 percent, compared to 27 percent and 22 percent, respectively, for non-Aboriginals); for the victims to know their killers; to involve alcohol or drug consumption on the part of both victims and accused persons; and, perhaps as a consequence, to involve victim precipitation (Brzozowski, Taylor-Butts, and Johnson 2006). This analysis suggests that Aboriginal homicide is different in a number of important respects from lethal violence among non-Aboriginal Canadians, which provides an empirical justification for the discrete examination of this particular group.

There is, however, one Canadian study that suggests that similar consideration ought to be extended to black Canadians. In an attempt to quantify the risk of homicide for blacks in Toronto, Canada’s largest city, Gartner and Thompson (2004) found that blacks—in particular, young black males—are disproportionately represented as homicide victims and offenders, and these researchers argue that a specific set of political, social, and economic conditions, past and present, have shaped the quantity and quality of this violence. Since at least the early 1990s, the homicide rate per 100,000 blacks in Toronto averaged 10.1 (compared to an overall rate of 2.4), and the risk of this violence has become increasingly concentrated in Toronto’s black community over time. Gartner and Thompson tie increases in black homicide over this period to a series of public policy decisions that undermined Toronto’s educational system, social welfare system, public health system, and the infrastructure and cohesion of disadvantaged communities.7 Taken together, they argue that these findings highlight the need for research that racially disaggregates homicide rates in order to better understand the causes and correlates of this violence but also to inform decisions about the types of policies required to reduce homicide in Toronto among those at the greatest risk. Unfortunately, the de facto ban on the dissemination of race-based data has made it incredibly difficult for researchers to document and examine racial/ethnic disparity in risk of lethal violence in Canadian cities.

As a consequence of this official silence and the suppression of data on issues related to race/ethnicity and crime, Toronto-based news media play a crucial role in constructing the issue of black homicide victimization and in shaping popular understandings of this violence (Henry and Tator 2000, 2002; Ezeonu 2008). Media coverage on homicide in Toronto—particularly since the early 1990s—portrays it as a phenomenon that overwhelmingly involves young black males killing other young black males with handguns (Porter and Campbell 2001; Shephard, Sheppard, and Davis 2002). High levels of so-called black-on-black violence are variously attributed to drug- and gang-related (p. 434) activity (Porter and Campbell 2001), the importation of a violent Jamaican subculture (Blatchford 2003), or rap culture that glamorizes and encourages the perpetration of a “thug” lifestyle (Rivers 2005; Wente 2005).

The news media has also increasingly acknowledged the importance of “place” in shaping the risk of homicide in Toronto, identifying a small number of “at-risk” neighborhoods that are thought to be rife “with the festering conditions that spawn and nurture violent crime” (James 2005). The discourse of pathology that characterizes much media discourse over poor, racialized urban neighborhoods in Toronto is in keeping with research that has long demonstrated that popular discourses over disadvantaged urban neighborhoods can produce deeply stigmatized and racialized spaces—thereby reproducing a stereotypical link between race, space, and serious violent crime in the public imagination (Purdy 2003; Wacquant 2007).

In a broader political climate that precludes an empirical examination of these issues, media constructions of violence involving the black community appear to be, for most Torontonians with little to no first-hand knowledge, an accurate depiction of reality. This is a problem, as news media coverage in Toronto, as elsewhere, has been criticized for constructing and shaping lay criminologies in public discourse over violent crime. In other words, studies demonstrate that media portrayals often simplify important and complex social issues into smaller problems that are viewed as the responsibility of individual people or the broader communities of which they are a part—rather than the result of flaws in the larger social structure (Henry and Tator 2000, 2002; Ezeonu 2008). Research has also suggested that public policy related to violence prevention and control is itself often informed by media portrayals of violent crime (Henry and Tator 2000; Ezeonu 2008). This is particularly worrisome in light of the fact that there are few empirical “checks” with which to balance media coverage of social and spatial concentrations in the risk of homicide in Canada. In what follows, I offer the results of my research on black homicide victimization in Toronto, along with directions for future research and policy implications, as a first step in this regard.

II. Case Study: The Social and Spatial Distribution of Black Homicide Victimization in Toronto, 1988–2003

There are two main ways that researchers in Canada, in theory, can gain access to homicide data: (1) through the Canadian Centre for Justice Statistics (CCJS), a division of Statistics Canada, and (2) from individual police departments and coroner’s offices.8

Police services in Canada are legally mandated to provide information on all homicides that occur to Statistics Canada, via the CCJS. When a homicide becomes known to police, the lead investigating officer completes a “Homicide Survey,” which provides detailed information on the homicide event, the victim, and accused person(s), (p. 435) if known. This information is then forwarded to the CCJS, where it is verified and catalogued in a national database. These data are then publicly disseminated via the production of reports and summary tables that document select aggregate (i.e. national-, provincial-, and territorial-level) patterns, trends, and characteristics of homicide offences, victims, and offender(s), as well as city-level counts and rates for select (usually recent) time periods. Raw data are not publicly available, nor are data for aggregations below the city-level. In sum, information on homicide in Canada that is produced for public consumption by the CCJS really provides little more than a broad-brush picture of lethal violence over short periods of time.9

Even if raw data on homicide in Canada were to be made publicly available, information on the race/ethnicity of the victims and offenders would, for the most part, be unavailable. Specific information on the racial/ethnic background of victims and offenders is only collected if they are of “Aboriginal origin” (i.e. Inuit, Metis, and First Nations) due to limited permission in a political and organizational culture that is otherwise not open to discussions of race, particularly if they also involve criminal victimization and offending. Further, the reliability of data that are gathered for Aboriginal groups is itself questionable; many police agencies in Canada refuse to report this information to Statistics Canada (Millar and Owusu-Bempah 2011). As such, although race-based data are routinely collected by police and other criminal justice agencies in Canada, they are not subsequently publicly or systematically disseminated.

More detailed information on homicide in Canada can be gathered from police and coroner’s files. The process to gain access to these files is, however, lengthy, and the success of such requests uncertain at best. In Canada, specific data requests must first be submitted, in writing, to the requisite public institution (the individual police organization, coroner’s office, etc.). Should requests submitted through such “normal access procedures” not be successful (due to, for example, organizational concerns over a variety of legal and privacy issues that often accompany the collection of “sensitive” data), researchers are required to file a Freedom of Information (FOI) request. Governed by the Freedom of Information and Protection of Privacy Act, this legislation provides individuals with a right of access to records and information under the custody or control of government ministries and agencies, including law enforcement. Yet this is not a straightforward process. In Ontario, for example, if the data requested from law enforcement agencies fall within the boundaries of a series of mandatory and discretionary exclusion categories (that again pertain largely to privacy issues), disclosure of these data can be (and often is) refused. If, on the other hand, a research request is ultimately successful, the researcher(s) will then enter into a legal agreement with that public institution, which will specify the terms and conditions under which these data can be collected.

A. Homicide Data Collection for This Research

In early 2004, I began the process of requesting, from the Toronto Police Service and the Office of the Chief Coroner of Ontario, files on all homicides in Toronto for the (p. 436) period 1988–2003. The Toronto Police Service requested that I file a formal FOI request, and 3 years later, when the process was complete, I was permitted to begin data collection in the homicide squad’s offices, with a research agreement that stipulated I not collect information on the racial/ethnic background of homicide victims and offenders and that I not share the data I collect with researchers outside of the immediate project. The Office of the Chief Coroner of Ontario, by contrast, processed my research request in-house (i.e., a more formal FOI request was not required), and, approximately 2.5 years after the initial request was made (and due to the length of time required to process that request), I began data collection at their offices in Toronto. The research agreement entered into with the Office of the Chief Coroner also expressly prohibited the collection of race-based information from their files.10

Data collection at the Toronto Police Service was on a full-time basis and took approximately 4 months; I worked out of the Chief Coroner’s Office for about 1 year on a part-time basis. From the files of both organizations, and for each case, I collected demographic information on the victim and accused person(s) (including sex, age, employment status, job type, drug/alcohol use at the time of the killing, prior criminal record, number of children, marital status, etc.), situational characteristics of the homicide (including the motive, method, address of and circumstances surrounding the killing, weapon type, etc.), and case disposition outcomes, where available. I also wrote a short narrative that summarized the incident description that was included in these reports.

A crime reporter at the Toronto Star provided me with a list of all homicides that had occurred in Toronto over the period of examination, including the name of the victim, the date, and location of the killing. Using an electronic newspaper search engine, I collected comparable information on victims, offenders, and the homicide incident, and wrote a short summary of the coverage. The newspaper coverage also enabled me to collect data on the racial/ethnic background of victims and, where possible, offenders; I relied on newspaper reports and photographs of victims and accused persons that were printed as part of those reports, assigning persons to an admittedly crude and overaggregated “black/nonblack” classification scheme. I was, however, able to cross-check my racial classifications against a list of black homicide victims over much of the period of investigation that was compiled by Toronto’s Black Action Defense Committee, preferring their classifications in the two cases in which there was a discrepancy. It is impossible to overemphasize my acknowledgment that assigning race in this way is fraught with the potential for misclassification and missing data, but, given institutional policies in Canada that ban the release of race-based statistics, there was no other alternative.

B. Data and Measures

The homicide data collected from these three sources were pooled over a 16-year period of examination in order to have a sufficient number of homicides with which to perform multivariate analyses.11 From this larger dataset, I selected out killings involving black victims (n = 225). Table 15.1 provides descriptive statistics for selected victim, incident, (p. 437) (p. 438) and offender characteristics of black homicides (n = 225) in Toronto’s neighborhoods between 1988 and 2003.

Table 15.1 Characteristics of homicides involving black victims in Toronto, 1988–2003 (n = 225)

Victim Sex (n = 225)

 % Male


 % Female


Victim Age (n = 979)

 Mean Victim Age 27

 % 1–15


 % 16–24


 % 25–34


 % 35–44


 % 45+


Victim Marital Status (n = 195)

 % Single (never married)


 % Married (including common-law)


 % Separated/Divorced (including common-law)


 % Other1


Victim Employment Status (n = 200)

 % Employed


 % Unemployed


 % Student


 % Other


Victim-Offender Relationship (n = 136)

 % Friends/Acquaintances


 % Strangers


 % Intimate partners


 % Illegal relationship


 % Other2


Offender Sex (n = 150)

 % Male


 % Female


Offender Age (n = 141)

 % 0–15


 % 16–24


 % 25–34


 % 35–44


 % 45+


Offender Race (n = 10)

 % Black


 % White


 % Other


Location of Killing (n = 225)

 % Residence


 % Public


 % Stores, bars


 % Semi-public


 % Car


 % Other


Method of Killing (n = 225)

 % Shot


 % Stabbed


 % Beaten


 % Strangled/Suffocated


 % Other3


(1) This category includes widowed, living together for a short time (less than 1 month), or off-and-on for short periods.

(2) This category includes housemates/roommates, neighbors, legal business relationships, co-workers, lovers’ triangles, family, and foster children.

(3) This category includes death by poisoning, arson, drowning, thrown or pushed from height, scalding, neglect, hit by car, overdose, and unspecified means.

Over this period, the majority of black homicide victims (83 percent) were males and young people. For example, 47 percent of victims were under the age of 24, with an additional 39 percent in the 25–34 age category. In other words, approximately 90 percent of all black homicide victims were under the age of 34 (the mean age was 27), which makes them younger than victims of homicide in Toronto more generally12 and mirrors the marked concentrations by age and sex in the American context. A large proportion of black victims were also unmarried and unemployed.

My dataset also contains information on a variety of incident characteristics, including the relationship between victim and offender (when known), the location of the homicide, and the method of and apparent motivation for the killing. Approximately one-third (34 percent) of black homicides involved friends/acquaintances (compared to 25 percent of total homicides in Toronto over the period of examination), 27 percent (p. 439) involved strangers (vs. 21 percent of total homicides), 13 percent involved male or female intimate partners (vs. 19 percent), and 10 percent involved illegal relationships (vs. 8 percent). In terms of location, 34 percent of black homicides occurred in a private residence (compared to 48 percent of total homicides in Toronto); 29 percent in public spaces (vs. 21 percent)—for example in streets, parks, or parking lots; 19 percent of black homicides occurred in stores and places of leisure (vs. 15 percent), such as bars and restaurants; and 7 percent of victims were killed in a vehicle (vs. 4 percent).The use of firearms figures prominently in black homicides; 64 percent of these killings involved a gun, compared with 35 percent of total homicides over the period of examination.

My ability to document characteristics of black homicide offenders is greatly constrained by missing data. This is, in part, a function of the fact that, for 33 percent of these cases in my dataset, no information on the offender was available. As such, the following statistics should be interpreted with an understanding that a considerable amount of data is missing for offenders. Perpetrators of homicides involving black victims in Toronto were overwhelmingly young black males. Of the cases in which an offender was identified, almost half were people under the age of 24 (the mean age was again 27), 95 percent were male, and 92 percent were black. This is consistent with the larger literature on homicide, which shows that the characteristics of homicide offenders are typically very similar to those of homicide victims and that most homicides are intraracial (Silverman and Kennedy 2004). In other words, as in the American context, it is the case that young black males tend to be killed by other young black males in Toronto. I now turn to an examination of the geography of this violence.

III. The Spatial Distribution of Black Homicide Victimization in Toronto

Each of the 965 homicides that occurred in Toronto between 1988 and 2003 was geocoded to one of 140 “neighborhoods,” based on the location of the killing, because I am interested in neighborhood contexts that generate violence events, not violent people.13 Again, I selected out those homicides involving black victims for analysis (n = 225). Figure 15.1 provides a geographic representation of the distribution of homicides involving black victims across Toronto’s neighborhoods between 1988 and 2003 (n = 225).14

Case StudyBlack Homicide Victimization in Toronto, Ontario, CanadaClick to view larger

Figure 15.1 Homicide in Toronto’s neighborhoods involving black victims, 1988–2003.

Over this period, 111 neighborhoods (79 percent) experienced a black homicide rate that fell between 0 and 1.2 per 100,000 population (59 of which experienced no incidents of this violence); 20 neighborhoods (14 percent) experienced a rate of 1.2–2.4, five neighborhoods (4 percent) experienced a rate of 2.41–3.6, and four neighborhoods (3 percent) experienced a rate of 3.61 to 6.02.This is consistent with U.S.-based research that has found that high levels of African-American homicide tend to be concentrated in a small number neighborhoods located in the inner-city. In Toronto, black homicide is similarly (p. 440) concentrated in a small number of neighborhoods, although the majority of neighborhoods that experience the highest rates of black homicide (i.e., 3.6–6.2 per 100,000) are located on the western fringe of the city. These neighborhoods, dubbed “inner cities on the outer edges” (Wente 2004), are illustrative of a trend toward the increased concentration of poverty in neighborhoods that ring the city center (United Way of Greater Toronto and Canadian Council on Social Development 2004; Hulchanski 2006), of which one consequence appears to be high levels of lethal violence.

IV. Measuring Neighborhood Context for Black Homicide Victimization in Toronto

My independent variables encompass the “usual suspects” that have had the greatest theoretical and empirical significance in the broader literature on neighborhoods and homicide. Because my period of examination spans 16 years, I combined measures of neighborhood characteristics in Toronto from four censuses (1986, 1991, 1996, and 2001) and averaged them to produce one score for each of the independent variables. To measure socioeconomic disadvantage, I use five variables often included in the (p. 441) U.S.-based literature on neighborhoods and homicide, either singly or as a composite index: median family income, the percentage of the neighborhood’s total income that was composed of government transfer payments, the percentage of neighborhood residents defined as low income by Statistics Canada, the percentage of neighborhood residents aged 15 and older who were unemployed, and the percentage of households in the neighborhood headed by either a male or female lone parent.15 Other measures of neighborhood characteristics included in the analyses are the percentage of neighborhood residents who identified their ethnic origin as black,16 the percentage of neighborhood residents aged 5 and older who had changed residences in the past 5 years, the percentage of dwellings in the neighborhood lived in by their owners,17 the percentage of neighborhood residents who immigrated to Canada within the last 10 years, and the percentage of neighborhood residents aged 15–24. I also include the logged population of the neighborhood as an offset variable.

Table 15.2 Characteristics of 140 neighborhoods in Toronto, averaged across four censuses (1986, 1991, 1996, 2001)




Standard Deviation

% Black





% Low Income





% Young Males





% Movers





% Unemployed





Avg. Hhold Income





% Lone Parents





% Gvt. Transfers





% Owners





% Rct. Immigrants





Average Population (nonlogged)





Table 15.2 presents descriptive data on my independent variables and demonstrates that there is substantial variability among Toronto’s 140 neighborhoods on the social, economic, and demographic characteristics I have measured. For example, only 1.4 percent of residents in one neighborhood owned their own homes, compared with 95 percent in another; the percentage of families headed by a lone parent ranged from 7 to 45 percent; and the average household income in Toronto’s poorest neighborhood was $22,637, compared with $223,232 in the wealthiest. As such, we can see that Toronto, the self-proclaimed “city of neighborhoods” also appears to be a city of great disparity across them.

In Table 15.3, I report the bivariate correlations between my measures of neighborhood characteristics and the number of black homicides in each neighborhood. Black (p. 442) homicide victimization is significantly and moderately/strongly associated with the proportion of black residents (.66), recent immigrants (.40), and the level of disadvantage in the neighborhood (.44),18 and significantly, but only weakly correlated with the proportion of young residents (.26), and the residential stability index (−.22). Thus, neighborhoods with higher levels of economic disadvantage and larger numbers of young, black, and recent immigrant residents tend to experience higher levels of black homicide victimization, whereas those characterized by residential stability typically experience lower levels of this violence. I next examine whether the relationships observed in the bivariate analysis are mirrored in the multivariate analysis.

Table 15.3 Pearson correlation coefficients among neighborhood characteristics and homicides with black victims in Toronto (n = 140), 1988–2003

Disadvantage Index


% Black


Residential Stability Index


% Young Males


% Recent Immigrants


(**) p < .01

V. “Neighborhood Effects” and Black Homicide Victimization in Toronto: Multivariate Analysis and Results

In this analysis, I use negative binomial regression, the most appropriate model for my data, which are overdispersed. An important consideration when analyzing spatial data is to measure and assess the degree of spatial autocorrelation, or dependency, among observations in a geographic space. The program ArcGIS 902 was used to carry out the Moran’s I test for spatial autocorrelation. The results indicate the presence of spatial autocorrelation; I therefore add a spatial lag to the regression model to capture the effect of neighboring locations.

The results of my multivariate analyses, presented in Table 15.4, indicate that the percentage of neighborhood residents who are young (15–24) and black are positively and significantly associated with black homicide victimization; none of the other measures (p. 443) of neighborhood characteristics is significantly associated with black homicide at the multivariate level. I now turn to a discussion of possible explanations for these results.

Table 15.4 Negative binomial regressions: neighborhood characteristics and homicides with black victims (n = 140)

Disadvantage Index

.039 (.054)

% Black

.177** (.040)

Residential Stability Index

.008 (.008)

% Young People 15–24

.159 (.071)

% Recent Immigrants

.057 (.066)

Spatial Autocorrelation Coefficient

−.018 (.114)



Log Likelihood


(***) = p < .001

(**) = p < .01

(*) = p < .05

Note: Standard errors in parentheses. Unstandardized parameter estimates presented.

VI. Discussion

Perhaps ironically in a country that is uncomfortable with official discussions of “race” and crime, only one of my measures of neighborhood characteristics—percent black residents—is associated with black homicide victimization in Toronto. This finding differs from a number of American studies that show that percent black is associated with homicide at the bivariate level, but not when measures of poverty are included in the multivariate model (Messner and Tardiff 1985; Sampson 1985), and begs the question why.

One answer is that a compositional effect is in operation. In other words, in neighborhoods with a preponderance of black residents, one would expect that a larger proportion of victims would also be black. Yet there might also be contextual effects at play that may help to explain why there is an association between the proportion of black residents in a neighborhood and black homicide victimization. A number of studies suggest that high rates of crime and violence in disadvantaged neighborhoods may be explained with reference to the emergence of local cultural adaptations that privilege violence as a means of status attainment or conflict resolution (Horowitz 1983; Fagan and Wilkinson 1998; Anderson 1999). These social norms that develop in response (p. 444) to structural disadvantage are especially applicable for understanding violent crime among young, racialized males who, it is argued, are more likely to internalize and act upon the dictates of these “street codes” in their neighborhoods. It may be the case that such codes are operating in Toronto’s neighborhoods and that they represent an intervening mechanism that may give rise to higher risks of homicide among black Torontonians.

A second and complementary explanation of the relationship between my percent black measure and lethal violence in Toronto’s neighborhoods involves a consideration of the experience of being poor and black in Canadian society. Studies demonstrate that a high proportion of black youth perceive that Canadian society generally and the Canadian criminal justice system in particular discriminate against them (Ruck and Wortley 2002). In response to this perceived discrimination, some may develop a set of cultural values or norms that justify or excuse the use of violence. To the extent that perceptions of injustice in Canadian society have created a sense of social and legal inequality among young black males in Toronto, violence may also become part of a repertoire of behaviors for coping with the problems in their lives.

The perception that one’s racial or ethnic group faces discrimination in the school setting and later in the labor market may also prompt some to seek alternative, and illegal, sources of income. Drug markets, which are often controlled by local gangs, may offer such an opportunity to people who feel that they have few to no options in the legitimate sphere. If it is the case that, due to perceptions of inequality in the opportunities available to them, some young black males in Toronto are more likely to become involved in gang- and drug-related activity than are young men from other racial/ethnic groups, levels of violence, including homicide, among this group would necessarily be higher.

In sum, then, cultural values that emerge in response to perceptions of injustice in Canadian society may operate in tandem with the rational choice to carry a weapon for the purposes of protection and/or status attainment, and/or to seek alternative sources of income in the illegal economy. These so-called cultural adaptations to structural inequalities may be useful for understanding why neighborhood racial composition seems to matter for understanding high levels of black homicide in Toronto’s neighborhoods. As will be discussed next, additional research into the microlevel processes that give rise to higher levels of lethal violence among black Torontonians would also be helpful in interpreting my results.

VII. Policy Implications

Perhaps the most obvious research-related policy implication that stems from this study pertains to the limited access that researchers in Canada have to homicide data and data on the racial/ethnic background of homicide victims and offenders. As previously discussed, these data are collected and catalogued in a number of official capacities, but (p. 445) they are not subsequently made available to researchers for analysis. Such policies hinder the development of theory and research on homicide that is more cross-national in scope. So, too, does the absence of data that would allow researchers to document racial disparity in violent victimization. Although criminal justice agencies in Canada routinely collect and internally use race/ethnicity-based data, these data are not made publicly available. Yet these data are crucial to measuring the existence and extent of racial/ethnic disparity in violent victimization and in establishing measures against which the efficacy of programs and policies designed to address the problem may be assessed. I therefore reiterate the important caveat that, in order to be effective, violence reduction policies must be implemented with an empirical understanding of the problems they are designed to prevent or control. This necessarily requires that researchers in Canada be granted access to official data on violent victimization—including that which is disaggregated by race/ethnicity.

A second policy implication that stems from this study lies in its utility to assist in the identification of neighborhoods vulnerable to high levels of lethal violence, for anticipatory intervention initiatives. However, one of the most difficult challenges for practitioners and policy makers interested in reducing levels of black homicide victimization in Toronto may lie in determining what not to do (Doob 2004). A common intervention in high-crime urban neighborhoods involves intensive “targeted” law enforcement strategies. Research, however, suggests that although these interventions may sometimes have positive short-term effects, they generally do not reduce crime and violence in the long term (Cohen, Gorr, and Singh 2003; Rosenfeld, Fornango, and Baumer 2005). Studies also suggest that in neighborhoods where relationships between residents and the police are already strained, aggressive policing strategies may not only further undermine those relationships but may also serve to increase local levels of violent crime (Walker 1992). Residents of disadvantaged urban neighborhoods typically articulate that the police are overly aggressive in their interactions with local residents and with their use of coercive power more generally, which can exacerbate alienation and conflict between the two groups. As a consequence, neighborhood residents may choose to bypass agents of the formal system altogether, relying instead on informal methods to redress interpersonal disputes (Anderson 1999; Kubrin and Weitzer 2003; Kane 2005). In addition, studies suggest that decreases in perceived police legitimacy may be associated with reactions of defiance and reduced compliance with the law (Tyler 1990). As such, increased levels of violent crime in some disadvantaged neighborhoods may stem, in part, from enforcement activities that undermine the police as a legitimate institution of formal social control (Tyler and Wakslak 2004; Kane 2005). Given the evidence suggesting that police crackdowns and related tactics are unlikely to have long-term beneficial effects, it would appear that caution should be exercised in assuming that no harm could come from such activities.

On the other hand, studies have shown that concentrating more general public health interventions in urban neighborhoods—for example, building neighborhood-level social and economic capital, increasing levels of community cohesion, and promoting collective action among residents—may have important declining effects on local levels (p. 446) of crime and violence (Bursik and Grasmick 1993; Sampson 1997). As such, it may be that more general policies that address social and structural deficits hold the most promise for a long-term reduction in homicide in Toronto’s black community. For example, school-based programs that encourage youth to become involved in and committed to school have a number of beneficial effects, including a reduction in levels of violent offending among “high-risk” youth (Smith, Lizotte, and Thornberry 1995). Social programs designed to improve pre- and postnatal care, enhance parenting skills, and otherwise promote healthy children and families can also have important crime prevention effects (Howell and Hawkins 1998; Tremblay and Japel 2003). In the long run, such programming may also be more cost effective than heavy-handed law enforcement interventions aimed solely at preventing and reducing crime and violence. This is because crime prevention and reduction initiatives, which typically fail to address the structural “root causes” of crime, tend to focus on one possible benefit: an immediate reduction in crime. Focused public health interventions, however, have the potential to achieve multiple benefits, including but not limited to a reduction in local levels of crime and violence over the long term.

VIII. Directions for Future Research

The findings from this study offer a first step in understanding black homicide victimization outside of the American context. In so doing, the findings prompt both additional questions and new directions for research. In this final section, I briefly discuss two.

The first involves exploring the microlevel processes that give rise to high levels of black homicide in Toronto’s neighborhoods, which would require the greater incorporation of qualitative methodologies into social ecological research. For example, as discussed earlier, it may be the case that “street codes” are operating in Toronto’s neighborhoods and that they represent an intervening mechanism that give rises to higher rates of violent crime among young black males who, it is argued, are more likely to internalize and act on the dictates of these codes. It is, however, crucial that this not be assumed given the very different history and legacy of racism experienced in Canada that has given rise to the social and economic conditions within which poor, urban blacks live and the issues that face their communities. Rather, whether and the extent to which microlevel processes identified in the U.S.-based literature apply in the Canadian context must be empirically demonstrated.

Finally, for the most part, attempts to examine “racial” disparity in homicide victimization are confined to the North American context. This is because, with a small handful of European exceptions, only the United States collects and reports race/ethnicity-based criminal justice data, whereas in Canada these data are often collected but not publicly disseminated. In most European and Scandinavian nations, conversely, there is no concept of “race” in the administrative and criminal justice registers that are (p. 447) used to produce statistics. Instead, there are extensive records on immigration status and/or nationality, which themselves are not collected in the North American context. Thus, although it is obviously not possible to make cross-national comparisons of “racial” disparity in homicide victimization outside of the North American context or of this disparity by immigration status/nationality outside of the European context—there is much to be learned about disparity in the risk of this violence among disadvantaged minority groups (i.e., relative to majority groups) in Western nations more generally. In this respect, cross-national research holds great potential to provide valuable theoretical and empirical information to policy makers and academics interested in urban homicide, policy interventions, and disadvantaged populations in Canada and around the world.


Anderson, Elijah. 1999. The Code of the Street: Decency, Violence, and the Moral Life of the Inner City. New York: W.W. Norton and Company.Find this resource:

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

      Blatchford, Christie. 2003. “Fear, Death Rule ‘Doomstown.’” The National Post (January 17).Find this resource:

        Boatswain, Sharon J., and Richard N. Lalonde. 2000. “Social Identity and Preferred Ethnic/Racial Labels for Blacks in Canada.” Journal of Black Psychology 26(2): 216–234.Find this resource:

          Browning, Christopher R., Reginald A. Byron, Catherine A. Calder, Lauren J. Krivo, Mei-Po Kwan, Jae-Yong Lee, and Ruth D. Peterson. 2010. “Commercial Density, Residential Concentration, and Crime: Land Use Patterns and Violence in Neighborhood Context.” Journal of Research in Crime and Delinquency 47:329–357.Find this resource:

            Brzozowski, Jodi-Ann, Andrea Taylor-Butts, and Sara Johnson. 2006. “Victimization and Offending Among Aboriginal People of Canada.” Juristat 26:1–31.Find this resource:

              Bursik, Robert J., and Harold G. Grasmick. 1993. Neighbourhoods and Crime: The Dimensions of Effective Community Control. New York: Lexington Books.Find this resource:

                Cohen, Jacqualine, Wilpen Gorr, and Piyusha Singh. 2003. “Estimating Intervention Effects in Varying Risk Settings: Do Police Raids Reduce Illegal Drug Dealing at Nuisance Bars?” Criminology 41(2): 257–292.Find this resource:

                  Doob, Anthony. 2004. “Preventing Violent Offending in Youth.” In From Enforcement and Prevention to Civic Engagement: Research on Community Safety, edited by Bruce Kidd and Jim Phillips. Toronto: Centre of Criminology.Find this resource:

                    Ezeonu, Ifeyani. 2008. “The Media Construction of ‘Black on Black’ Violence in Toronto.” Africana: A Journal of Ideas on Africa and thee African Diaspora 2(1): 33–76.Find this resource:

                      Fagan, John, and Deanna Wilkinson. 1998. “Guns, Youth Violence, and Social Identity in Inner Cities.” In Crime and Justice: A Review of Research, vol. 24, edited by Michael Tonry and Mark H. Moore. Chicago: University of Chicago Press.Find this resource:

                        Farrington, David P. 2000. “Explaining and Preventing Crime, The Globalization of Knowledge: The American Society of Criminology Presidential Address.” Criminology 38(1): 1–24.Find this resource:

                          Galabuzi, Grace Edward. 2006. Canada’s Economic Apartheid: The Social Exclusion of Racialized Groups in the New Century. Toronto: Canadian Scholars Press.Find this resource:

                            Gartner, Rosemary, and Sara K. Thompson. 2004. “Trends in Homicide in Toronto.” In From Enforcement and Prevention to Civic Engagement: Research on Community Safety, edited by Bruce Kidd and Jim Phillips. Toronto: Centre of Criminology.Find this resource:

                              Ghee, Kenneth L. 1990. “The Psychological Importance of Self-Definition and Labeling: Black Versus African American.” Journal of Black Psychology 17:75–93.Find this resource:

                                Harrell, Erika. 2007. Black Victims of Violent Crime. Bureau of Justice Statistics Special Report Washington, DC: US Department of Justice: Office of Justice Programs.Find this resource:

                                  Henry, Frances, and Carol Tator. 2000. The Colour of Democracy: Racism in Canadian Society (3rd edition). Toronto: Nelson. (p. 451) Find this resource:

                                    ——. 2002. Discourses of Domination: Racial Bias in the Canadian English-language Press. Toronto: University of Toronto Press.Find this resource:

                                      Horowitz, Ruth. 1983. Honor and the American Dream: Culture and Identity in a Chicano Community. New Brunswick: Rutgers University Press.Find this resource:

                                        Howell, James, and J. David Hawkins. 1998. “Prevention of Youth Violence.” In Youth Violence, edited by Michael Tonry and Mark H. Moore. Vol. 24 of Crime and Justice, A Review of the Research, edited by Michael Tonry. Chicago: University of Chicago Press.Find this resource:

                                          Hulchanski, J. David. 2006. “The Suburbanization of the Non-Gentry: Toronto’s Inner Suburbs.

                                          James, Royson. 2005. “Hear Our Prayer.” Toronto Star (14 August).Find this resource:

                                            Jiwani, Yasmin. 2002. “Erasing Race: The Story of Reena Virk.” In Violence Against Women, New Canadian Perspectives, edited by Katherine M. J. McKenna and June Larkin. Toronto: Inanna Publications.Find this resource:

                                              Kane, Robert J. 2005. “Compromised Police Legitimacy as a Predictor of Violent Crime in Structurally Disadvantaged Communities.” Criminology 43(2): 469–498.Find this resource:

                                                Kubrin, Charis E. 2003. “Structural Covariates of Homicide Rates: Does Type of Homicide Matter?” Journal of Research in Crime and Delinquency 40: 139–170.Find this resource:

                                                  Kubrin, Charis E., and Tim Wadsworth. 2003. “Identifying the Structural Correlates of African-American Killings: What Can We Learn from Data Disaggregation?” Homicide Studies 7:3–35.Find this resource:

                                                    Kubrin, Charis E., and Ronald Weitzer. 2003. “Retaliatory Homicide: Concentrated Disadvantage and Neighbourhood Crime.” Social Problems 50(2): 157–180.Find this resource:

                                                      LaFree, Gary. 2007. “Expanding Criminology’s Domain: The American Society of Criminology 2006 Presidential Address.” Criminology 45(1): 1–31.Find this resource:

                                                        Messner, Steven, and Kenneth Tardiff. 1985. “The Social Ecology of Urban Homicide: An Application of the ‘Routine Activities’ Approach.” Criminology 23(2): 241–268.Find this resource:

                                                          Millar, Paul, and Akwase Owusu-Bempah. 2011. “Whitewashing Justice in Canada: Preventing Research through Data Suppression.” Canadian Journal of Law and Society 26(3): 653–661.Find this resource:

                                                            Mosher, Clayton. 1998. Discrimination and Denial: Systemic Racism in Ontario’s Legal and Criminal Justice Systems: 1892–1961. Toronto: University of Toronto Press.Find this resource:

                                                              Peterson, Ruth D., and Lauren J. Krivo. 2005. “Macrostructural Analyses of Race, Ethnicity, and Violent Crime: Recent Lessons and New Directions for Research.” Annual Review of Sociology 31:331–356.Find this resource:

                                                                Porter, M., and M. Campbell. 2001. “Three Neighbourhoods Grapple with Shootings.” Toronto Star (July 28).Find this resource:

                                                                  Purdy, Sean. 2003. “‘Ripped Off’ by the System: Housing Policy, Poverty, and Territorial Stigmatization in Regent Park Housing Project: 1951–1991.” Labour 52:1–46.Find this resource:

                                                                    Rivers, E. 2005. “The Sins of the Fathers are Visited on Black Youth.” Globe and Mail (December 2).Find this resource:

                                                                      Roberts, Julian, and Anthony Doob. 1997. “Race, Ethnicity, and Criminal Justice in Canada.” Crime and Justice 21:469–522.Find this resource:

                                                                        Rosenfeld, Richard, Timothy M. Bray, and Arlen Egley. 1999. “Facilitating Violence: A Comparison of Gang-Motivated, Gang-Affiliated, and Non-Gang Youth Homicides.” Journal of Quantitative Criminology 15(4): 495–516.Find this resource:

                                                                          Rosenfeld, Richard, Robert Fornango, and Eric Baumer. 2005. “Did Ceasefire, Comstat, and Exile Reduce Homicide?” Criminology and Public Policy 4(3): 419–450. (p. 452) Find this resource:

                                                                            Ross, Rupert. 1995. “Restorative Justice: Exploring the Aboriginal Paradigm.” Saskatchewan Law Review 59:431–435.Find this resource:

                                                                              Royal Commission on Aboriginal Peoples. 1995a. Bridging the Cultural Divide. Ottawa: Canada Communications Group.Find this resource:

                                                                                ——. 1995b. Looking Forward, Looking Back. Ottawa: Canada Communications Group.Find this resource:

                                                                                  Ruck, Martin D. and Scot Wortley. 2002. “Racial and Ethnic Minority High School Students’ Perceptions of School Disciplinary Practices: A Look at Some Canadian Findings.” Journal of Youth and Adolescence 31(3): 185–195.Find this resource:

                                                                                    Rudin, Jonathan. 2005. “Aboriginal Peoples and the Criminal Justice System: A Background Paper prepared for the Ipperwash Inquiry.

                                                                                    Sampson, Robert J. 1985. “Race and Criminal Violence: A Demographically Disaggregated Analysis of Urban Homicide.” Crime and Delinquency 31:47–82.Find this resource:

                                                                                      ——. 1987. “Urban Black Violence: The Effect of Male Joblessness and Family Disruption.” American Journal of Sociology 93(2): 348–382.Find this resource:

                                                                                        ——. 1997. “The Community.” In Crime, edited by James Q. Wilson and Joan Petersilia. San Francisco: ICS Press.Find this resource:

                                                                                          Shephard, A., S. Myers, J. Sheppard, and K. Davis. 2002. “Victims Black in 42% of Non-Domestic Killings: Analysis Shows Vast Majority of Suspects in Those Slayings also Black.” National Post (8 November).Find this resource:

                                                                                            Sigelman, Lee, Steven A. Tuch, and Jack K. Martin. 2005. “What’s in a Name?: Preference for ‘Black’ Versus ‘African American’ Among Americans of African Descent.” Public Opinion Quarterly 69(3): 429–438.Find this resource:

                                                                                              Silverman, Robert A., and Leslie W. Kennedy. 2004. Deadly Deeds: Murder in Canada Scarborough: Nelson.Find this resource:

                                                                                                Sinclair, Murray. 1994. “Aboriginal People, Justice, and the Law.” In Continuing Poundmaker and Reil’s Quest: Presentations Made at a Conference on Aboriginal Peoples and Justice, edited by Richard Gosse, James Youngblood Henderson, and Roger Carter. Saskatoon: Purich Publishing.Find this resource:

                                                                                                  Smith, Carolyn, Alan J. Lizotte, and Terence P. Thornberry. 1995. “Resilient Youth: Identifying Factors that Prevent High-Risk Youth From Engaging in Delinquency and Drug Use.” Current Perspectives on Aging and the Life Cycle 4: 217–247.Find this resource:

                                                                                                    Stamatel, Jean. 2009. “Contributions of Cross-National Research to Criminology at the Beginning of the 21st Century.” In Handbook on Crime and Deviance, edited by Marvin D. Krohn, Alan J. Lizotte, and Gina Penly Hall. New York: Springer.Find this resource:

                                                                                                      Tremblay, Richard E., and Christa Japel. 2003. “Prevention During Pregnancy, Infancy, and the Preschool Years.” In Early Prevention of Adult Antisocial Behaviour, edited by David P. Farrington and Jeremy W. Coid. New York: Cambridge University Press.Find this resource:

                                                                                                        Tyler, T. R., and Cheryl J. Wakslak. 2004. “Profiling and Police Legitimacy: Procedural Justice, Attributions of Motive, and Acceptance of Police Authority.” Criminology 42:243–281.Find this resource:

                                                                                                          Tyler, Tom R. 1990. Why People Obey the Law. New Haven, CT: Yale University Press.Find this resource:

                                                                                                            United Way of Greater Toronto and the Canadian Council on Social Development. 2004. “Poverty by Postal Code: The Social Geography of Neighborhood Poverty, 1981–2001 (p. 453)

                                                                                                            Wacquant, Loic. 2007. “Territorial Stigmatization in the Age of Advanced Marginality.” Thesis Eleven 91:66–77.Find this resource:

                                                                                                              Walker, Samuel. 1992. The Police in America: An Introduction. New York: McGraw-Hill.Find this resource:

                                                                                                                Wente, Margaret. 2004. “Inner Cities on the Outer Edge.” Globe and Mail (April 13).Find this resource:

                                                                                                                  ——. 2005. “Get Mad, We’re Bein’ Had, Gangsta Rap’s Really Bad.” Globe and Mail (December 1).Find this resource:

                                                                                                                    Xie, Min. 2010. “The Effects of Multiple Dimensions of Residential Segregation on Black and Hispanic Homicide Victimization.” Journal of Quantitative Criminology 26:237–268. (p. 454) Find this resource:


                                                                                                                      (1) . In this paper, I use the term “race” while acknowledging it is a socially constructed concept that does not exist in the world in any ontologically objective way. However, the concept of race carries with it very real social, economic, and political consequences, which means that it matters in a number of important respects, including in shaping the risk of violent victimization and offending.

                                                                                                                      (2) . Research on racial self-designation has demonstrated that the label “African American” is increasingly preferred among black Americans (Ghee 1990; Sigelman, Tuch, and Martin 2005), whereas in Canada, the current self-label preference is “black” or “black Canadian” (Boatswain and Lalonde 2000). It may be interesting to note, however, that many of my nonblack students refer to blacks in Canada as “African Americans,” which suggests that the politics of black identity in Canada are informed in important ways by the American context (Boatswain and Lalonde 2000).

                                                                                                                      (3) . There are reasons to expect that there would be both similarities and differences in the Canadian black experience when compared to that of African Americans. Reasons to expect similarities include porous borders, the U.S. cultural influence, and deeply embedded racial resentments and animus that are characteristic of both countries (although perhaps more pronounced in the United States), past and present. At the same time, lower levels of poverty concentration and residential segregation in Canada, combined with a larger and better quality public housing stock, universal health and unemployment insurance programs, and a variety of family support programs may have resulted in a better quality of life for low-income black Canadians relative to their American counterparts.

                                                                                                                      (4) . It should be noted that “race” data are seldom recorded by state agencies and less often reported. Canada is an outlier in this regard among English-speaking countries, but is in the mainstream of developed countries.

                                                                                                                      (5) . As Roberts and Doob (1997, pp. 489–90) argue, “unlike crimes involving other groups within Canadian society, there seems to have been little political concern within the Aboriginal community or within the criminal justice policy community about public discussion of apparently higher rates of Aboriginal crime.” This has led to the production of official statistical reports on criminal offending that “could not have been written about any group in Canada other than Aboriginal people.” In other words, although the collection and dissemination of racially disaggregated criminal justice data is routine with respect to Aboriginal Canadians, it is forbidden and widely regarded as politically incorrect for other racialized groups.

                                                                                                                      (6) . As Rudin (2005) argues, three explanations have typically been advanced as significant causal factors in the literature on Aboriginal overrepresentation in the Canadian criminal justice system: (1) culture clash, which emphasizes the discrepancies between Aboriginal and Western concepts of justice (Sinclair 1994; Ross 1995); (2) socioeconomic conditions that see Aboriginal Canadians living in conditions of extreme socioeconomic and political deprivation (Rudin 2005); and (3) the impact of colonialism (Royal Commission on Aboriginal Peoples 1995a, 1995b). These topics have been addressed in depth by Canadian scholars and are outside the scope of this essay, although interested readers should look to the works cited here for additional information.

                                                                                                                      (7) . There is a growing literature in Canada that uses qualitative methodologies to provide a more in-depth analysis of the interrelated social, political, and economic circumstances that may have prompted a set of negative outcomes, including higher rates of violent victimization and offending, among some segments of the black population in Toronto (see for example, Galabuzi 2006; Ezeonu 2008).

                                                                                                                      (8) . In Canada, there are ways that university-affiliated researchers may gain access to data for social science, health, and policy research. For example, the Toronto Region Statistics Canada Research Data Centre (RDC)—a research facility funded by three Ontario universities and the federal government—houses a series of core datasets available to researchers affiliated with the funding universities, access to which usually requires a formal application process and background check. Currently, the only data on violent crime that are available through the Toronto RDC are of the self-report variety, which is, for obvious reasons, not useful to researchers examining homicide victimization. In the early 2000s, there was some movement toward the implementation of a pilot project that would facilitate the release of data from the Homicide Survey to researchers via the RDC, although, to date, there is no evidence of any movement on that front.

                                                                                                                      (9) . In the United States, hospital admission data for intentional injury and vital statistics on deaths are used to test the reliability of police and victim data. In Canada, researchers can access the Canadian Vital Statistics Death Database to verify official criminal justice data, but, as is the case in Canada more generally, information on the racial/ethnic background of victims is not provided. As such, it was not possible for me to use this database to gain additional insight into black homicide victimization in Canada.

                                                                                                                      (10) . I did, however, have many off-the-record conversations with coroners and senior police officers who emphatically stated their belief that the current ban on the release of racially disaggregated criminal justice data in Canada is greatly hindering a proper examination of an important public health issue. But, given the institutional stalemate in this regard, they argued “no organization is prepared to be the first to blink.” The suppression of these data is thought by many to be rooted in efforts on the part of criminal justice agencies to avoid the controversy inherent in linking race and crime, a topic that tends to be a political hot potato in Canada. Some have even argued that the language of “high-crime” or “at-risk” neighborhoods that pervades official discourse on violent crime in Toronto is used as a metaphor for communicating notions of race, thereby using geographic space to allude to a very specific subset of the black community (usually young black males), which itself is inextricably linked to high levels of violence in the public imagination. This is consistent with Canadian research that demonstrates that media portrayals and official discourse communicates race and racism by “omission” or strategic absence (Jiwani 2002). Put simply, when it comes to official discourse over serious violent crime in Toronto, the concept of “race” is often the proverbial elephant in the room.

                                                                                                                      (11) . I chose a starting point for the analysis that made practical sense. That is, I required a time period that would include a sufficient number of homicides to perform analyses of total and disaggregated homicide counts in Toronto’s 140 neighborhoods and one that corresponded with several Canadian censuses from which to obtain and average information on neighborhood characteristics. The end point for this study, however, was determined solely by Toronto Police Service (i.e., I was granted access to their files up to and including December 31, 2003).

                                                                                                                      (12) . By comparison, over the period of examination, 59 percent of total homicides in Toronto involved victims under the age of 34, and the mean victim age was 34.

                                                                                                                      (13) . These neighborhood delineations used in this research were constructed by Toronto’s Social Policy and Research Unit, based on social service areas defined by main streets, former municipal boundaries, and/or natural and manmade boundaries, such as rivers or highways. They were also chosen as they reduced the city’s 531 census tracts to a more manageable 140 neighborhoods.

                                                                                                                      (14) . There are a number of standard classification methods used to map quantities—in this study I used the equal interval class break method, which divides the range of values into equal-sized subgroups. The major advantage of this classification method lies in its ease of interpretation; other methods (such as natural breaks) are not always easily understood by map users, and the class breaks (range intervals) may not be intuitive. I did, however, group the highest two classes together in a custom classification technique, due to the small number of cases in each class break.

                                                                                                                      (15) . Including each of these measures separately in my multivariate model could create problems due to multicollinearity. To deal with this, I conducted a principal components factor analysis on my measures of socioeconomic disadvantage and my measure of lone-parent families; these variables are not only highly associated, they also load on a single factor, which I label a “disadvantage index.”

                                                                                                                      (16) . The percent black measure, which figures prominently in prior research on neighborhood effects and lethal violence, was excluded from the aforementioned factor analysis despite its strong correlations with the five measures of economic disadvantage. My choice to retain the racial composition of Toronto’s neighborhoods as a separate indicator in my analyses stems from Rosenfeld et al.’s (1999, p. 502) assertion that “[r]‌ace is conceptually distinct from ‘disadvantage,’ and treating them as attributes of the same dimension confounds attempts at untangling their distinct influences on levels of violence in a community.”

                                                                                                                      (17) . In keeping with U.S.-based research (e.g., Kubrin 2003; Browning et al. 2010), I created an index of “residential stability” that combined measures of population turnover and home ownership.

                                                                                                                      (18) . In other analyses (not reported here), the five measures of economic disadvantage (percent low income, percent unemployed, average household income, percent lone parents, and percent government transfer payments) are all highly correlated with each other: their intercorrelations range from −.64 (median family income and lone-parent families) to.93 (low income and unemployed residents). Including each of these separately in my multivariate models could create problems with multicollineariety, so I conducted a principal components factor analysis; all measures of economic disadvantage load on a single factor, which I label an “economic disadvantage index.”