Information identified as archived is provided for reference, research or recordkeeping 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.
The research findings for Edmonton, Halifax and Thunder Bay show that crime is not distributed randomly in urban areas, but is instead concentrated in particular neighbourhoods, especially those in city centres. In Edmonton, Halifax and Thunder Bay, the spatial distribution of property crime is, despite some differences, strongly related to that of violent crimes.
Several differences between the characteristics of high-crime neighbourhoods and those of lower-crime neighbourhoods were noted. These differences can be grouped under three broad dimensions, demographic, socio-economic and functional.
In the cities studied, the demographic characteristics of high-crime neighbourhoods differ from those of lower-crime neighbourhoods. High-crime neighbourhoods have a higher density of the population at risk, and that population has specific characteristics. The population of high-crime neighbourhoods has a larger proportion of single people, people living alone, young males aged 15 to 24, Aboriginals, people who moved in the year preceding the census and lone-parent families.
The analyses presented here do not establish causal links between these residents and the crime level in their neighbourhood. However, many studies have found links between these demographic characteristics and higher rates of victimization and even of delinquency (Kong 2005). These characteristics also play a role in the ability of neighbourhood residents to exercise supervision and informal social control (Cohen and Felson 1979).
In the cities studied, crime appears to be higher in neighbourhoods whose residents have more limited access to socio-economic resources. High-crime neighbourhoods are characterized by a population that is more disadvantaged in economic terms (higher unemployment rate and proportion of low income and government transfers, and lower incomes), and they have a smaller proportion of highly educated people. A larger proportion of the population of high-crime neighbourhoods spend more than 30% of their income on shelter, and a smaller number of owners occupy their dwelling, regardless of whether these neighbourhoods are located near city centres or are on the periphery of the municipality. High-crime neighbourhoods are also characterized by older dwellings or dwellings in need of major repairs.
The overlapping of these different socio-economic characteristics in cities was abundantly shown by the poverty and social exclusion project of the Canadian government's Policy Research Initiative (Horizon 2004). The Research Initiative's conclusions emphasized that the phenomenon of social exclusion and the persistence of low income are closely related (Lock, Kunz and Frank 2004). Many of the factors associated with the persistence of low income reflect the lack, ineffectiveness or disruption of social networks, more especially the social ties that provide access to income from stable paid employment (Hatfield 2004). People belonging to groups at risk (lone-parent families, elderly people living alone, people with a disability that limits their ability to work, Aboriginals living off reserve and new immigrants) share a number of problems, but each stands out by a specific event, whether it be a change in family status or even the lack of family status, a health problem or a move (Hatfield 2004).
In this context, then, high neighbourhood crime rates appear to reflect the absence, disruption or ineffectiveness of social networks that enable people to participate in the community and exert social control. Crime would appear to be a symptom of social exclusion, with social exclusion in turn blocking neighbourhood residents from exerting social control.
The functional characteristics of neighbourhoods play a role in the variation in crime levels registered at the neighbourhood level. High-crime neighbourhoods are the busiest neighbourhoods, either because they are located near city centres or because they support intense levels of commercial activity.
While the city centres are the largest hot spots in Edmonton, Halifax and Thunder Bay in absolute numbers, some places have higher crime rates. Most of these hot spots are areas of intense commercial activity (shopping malls and megastores). These places are characterized by a relatively large number of property crimes (most of which are thefts under $5,000) but also, to a lesser extent, violent crimes. West Edmonton Mall is an especially obvious example of this type of hot spot.
As to residential neighbourhoods, they are moderate crime areas. Various types of crimes are recorded there, including breaking and entering and assault, but they do not exhibit any particular spatial concentration, except for a few very high-density housing developments.
Other hot spots are institutional: schools, universities, hospitals, etc. While these places have more of certain types of crimes (such as arson incidents in schools), the population at risk in these places is heavily underestimated in our analyses, and neither students nor patients are taken into account.
When all other study characteristics are taken into account, a limited number of factors are found to be linked to the variation in the crime rate at the neighbourhood level. The set of explanatory factors varies in a specific way according to the city studied and the type of crime, violent or property. The three major dimensions (demographic, socio-economic and land use) are included in the various explanatory models.
Whereas some variables do not contribute significantly to the multivariate regressive models, this does not invalidate the role that they may play in the organization of crime. Indeed, it is owing to the combined effect of a number of characteristics that some neighbourhoods are more at risk. Owing to the accumulation and overlapping of these characteristics, some neighbourhoods are especially at risk (Massey 1996, Forrest and Kearns 2001, Sampson et al. 2002).
The analyses conducted in this study bear out the notion that the spatial organization of crime must be understood as the result, at a given time, of a slow and complex process of urban development. Neighbourhoods evolve with their inhabitants. Buildings age and are renovated or fall into disrepair; residents move away or remain; and communities are displaced, rebuilt and transformed.
Neighbourhoods are configurations of physical and symbolic conditions that shape everyday experience and the identity of their inhabitants (Debarbieux 2003). These conditions, which are specifically local, can play a major role in the occurrence of crime. It is in this evolving and complex context, situated in a specific time and place, that crime, its spatial organization and its links to the neighbourhood, must be understood.
These results suggest that the development of crime reduction strategies could be based on the local specifics of the demographic, socio-economic and land-use dimensions. If crime reduction strategies are based on the specific needs of each city—i.e., its history and the means available at the neighbourhood and community level—they will be more likely to achieve their objectives.
This multitude of demographic, socio-economic and land-use characteristics also suggests that a range of stakeholders should be involved and partnerships formed among different local players1 when developing and implementing crime prevention strategies. In fact, from the perspective of long-term change and the development of well-being in Canadian communities, crime prevention efforts should focus on creating an environment conducive to the broad and effective participation of partners in crime prevention, at all levels.
Studies based on the spatial analysis of crime should include a mass of information sources to adequately cover crime's different dynamics and aspects. The studies in this report were conducted using statistical data from police services, which include only crimes that are reported to them and confirmed by a police investigation. Many factors can influence police-reported crime rates, including the willingness of the public to report crimes to the police, and changes in legislation, policies or enforcement practices.
Thus, in the coming years it would be useful to examine, at the neighbourhood level, the information collected in victim and offender surveys, which in turn would provide a picture conducive to developing new crime prevention strategies. Surveys of the population would identify and better define the role of community involvement and social inclusion and exclusion in combating crime. Sampson et al. (2002) examined the theoretical and technical difficulties in measuring a number of neighbourhood characteristics relevant to crime issues. In particular, they mention the value of measuring social networks, norms and collective efficacy, institutional resources, community involvement and the spatial routine of the residents and users of neighbourhoods.
Understanding the factors related to change over the years can help policy makers develop crime prevention and reduction strategies, and evaluate existing programs. With the arrival of the 2006 Census data, it will be possible to obtain new demographic and socio-economic data at the neighbourhood level. These data will offer the opportunity to focus on the change over time and thus, for the first time, allows us to look at the factors associated with variations in crime at the neighbourhood level, and see how these evolve together, using Geographic Information System (GIS) technology.
The explanatory models and research results presented here do not take into account the activities of criminal gangs or the interactions between them. These activities and interactions, which for the most part are intrinsically territorial in nature, undoubtedly affect variations in crime levels at the neighbourhood level. Currently, the effect of these activities and interactions is, to some extent, taken into account by the use of spatial lag variables in autoregressive models, but no data enable us to determine what proportion of the spatial variables is attributable to them. The data needed to quantify interactions between criminal gangs are not available through the Uniform Crime Reporting (UCR) Survey. However, using the most recent version of the UCR Survey (2005), it will be possible to identify the presence of activities related to organized crime in neighbourhoods.
Also, despite the operational problems that this poses, incorporating existing social programs and crime reduction programs into subsequent analyses might make it possible to get a better understanding of the effectiveness of those programs with respect to crime prevention. It is important to pursue research on how to develop and evaluate strategies such as after-school programs and Neighbourhood Watch. This information would assist in answering questions such as the following: Do crime reduction programs actually reduce criminal activity? Does criminal activity merely shift to other neighbourhoods? What types of social or crime prevention programs are the most effective?
- Including partnerships among government departments, community groups, non-government organizations, the business community and citizens.
- Date modified: