An international perspective on criminal victimisation 

View the most recent version.

by Julie Sauvé and Kwing Hung

Results of the International Crime Victimization Survey 2004-2005 
According to the ICVS, prevalence rates of victimization in Canada differed little from international averages
Canadians reported a lower proportion of victimization incidents to the police compared to the average of the 30 countries participating in the ICVS
Summary
Methodology
Data limitations
Tables
Reference
Notes

Results of the International Crime Victimization Survey 2004-2005 

Adaptation of the publication Criminal Victimisation in International Perspective — Key findings from the 2004-2005 ICVS and EU ICS¹

Data on crime in Canada are drawn from two main sources: police-reported surveys and self-reported victimization surveys. Police-reported surveys capture criminal incidents that have come to the attention of the police as well as those that have been detected through police investigation. Police-reported data are drawn from the Uniform Crime Reporting (UCR) Survey and have been submitted to Statistics Canada each year since 1962. Victimization surveys collect information on respondents' personal accounts of criminal victimization and whether these incidents were reported to the police. Canada's national victimization survey is conducted every 5 years by Statistics Canada through the General Social Survey (GSS)². These two data sources are complementary tools for determining the nature and extent of crime and victimization in Canada.

Canada also participates in the International Crime Victimization Survey (ICVS). The most recent cycle of the survey was conducted in 2004 by Léger Marketing through the Department of Justice of Canada. The survey collected self-reported information from residents of 30 countries on their experiences of criminal victimization³. Through the use of a standard questionnaire, survey procedures and techniques for producing the results, the ICVS has been able to address the challenges in comparing victimization data from various countries.

While data from the international survey are not directly comparable with data from Canada's national survey of victimization (GSS) given different definitions of criminal offences, methodology and samples, these data allow for a comparisons of Canada's results to those of other countries who participated in the 2004-2005 ICVS survey⁴. The purpose of this Juristat article is to highlight those results, focusing on victimization prevalence rates⁵, rates of reporting to the police and respondents' satisfaction with the police work.⁶

According to the ICVS, prevalence rates of victimization in Canada differed little from international averages

The international survey measures the prevalence of victimization among people aged 16 and over, based on a series of ten offences (Text box 1). For all offences, victimization prevalence rates in Canada were similar to the international averages of the 30 countries that participated in the ICVS.

According to the 2004⁷ survey, 17% of Canadians aged 16 and over had been victims of at least one crime measured by the ICVS during the year preceding the survey. This rate was similar to the overall international victimization rate (16%) (Chart 1).

Victimization prevalence rates for 2004-2005 varied from one country to another, with Spain, Japan, Hungary and Portugal registering the lowest victimization rates (between 9% and 10%). In contrast, Ireland, England & Wales and New Zealand were among the countries with the highest overall victimization rates. 

Chart 1
According to the ICVS, Canada's victimization rate was similar to the average of the 30 participating countries, 2004-2005

Notes: Based on responses from victims of at least one out of the 10 types of offences in the year preceding the survey. In general, a difference less than 2.5 percentage points between two countries is not statistically significant.
1. Questions on sexual assault incidents were excluded from the survey in Australia. Thus, their national rate is estimated to be 1% lower.
Source: Van Dijk, Jan, John Van Kesteren, Paul Smit, Tilburg University, UNICRI and UNODC, (2007) 'Criminal Victimisation in International Perspective: Key Findings from the 2004-2005 ICVS and EU ICS'.

For almost all countries participating in the ICVS, the offences with the highest victimization rates were theft of personal property, theft from a car and theft of a bicycle. In Canada, the highest victimization rate was for theft from a vehicle. Nearly 5% of persons aged 16 and over had property taken from their vehicle.

Canadians reported a lower proportion of victimization incidents to the police compared to the average of the 30 countries participating in the ICVS

Not all incidents of criminal victimization are reported to the police. Across all participating countries, slightly more than half the population (53%) reported a victimization incident to the police. This rate is based on the reporting of five offence types comprised of theft from a car, theft of a bicycle, burglary, attempted burglary and theft of personal property. According to the ICVS, Canada, along with Finland and Luxembourg, ranked relatively low with respect to the reporting of one of those incidents to the police (Chart 2). With a rate below the international average, 48% of Canadian victims of such crime incidents in 2004 reported the incident to the police. Austria and Belgium had the highest reporting rates (70% and 68% respectively) on the international scale. Victims in Mexico were much less likely to report their victimization incidents to the police compared to all other countries surveyed (16%). The next lowest rate of reporting to the police was in Iceland (40%). 

Chart 2
Compared to other countries participating in the ICVS, less victimization incidents were reported to the police in Canada, 2004-2005

Notes: Based on responses from victims of at least one out of the 5 types of offences (theft from a car, theft of a bicycle, burglary, attempted burglary and theft of personal property) in the year preceding the survey. In general, a difference less than 4 percentage points between two countries is not statistically significant.
Source: Van Dijk, Jan, John Van Kesteren, Paul Smit, Tilburg University, UNICRI and UNODC, (2007) 'Criminal Victimisation in International Perspective: Key Findings from the 2004-2005 ICVS and EU ICS'.

According to the results of the ICVS, though Canadians reported a lower proportion of incidents to the police, when they did report, they were satisfied with the police response. In cases where Canadians reported theft from a car, burglary, robbery, sexual offences or assault, two-thirds reported that they were satisfied with how the police responded. Additionally, in 2004, 86% of Canadians aged 16 and over believed in 2004 that the police were doing a good or excellent job at controlling crime in their area. Canada, along with Finland and the United States, were among the participating countries in the ICVS whose population was the most satisfied with the job done by the police.

Summary

Canada did not stand out from other countries participating in the ICVS - criminal victimization rates were very close to the international averages for the 30 participating countries in the ICVS. Like the populations of other countries, Canadians were mainly victims of crimes against property. The reporting rates to the police of incidents in which Canadians were a victim were below the international average. However, the findings show that Canadians have a positive opinion regarding the job done by the police with respect to crime.

While Canada has participated in every cycle of the ICVS, there is also a national survey of victimization, conducted by Statistics Canada every five years since 1988. The General Social Survey on victimization, carried out on a larger scale than the ICVS, draws on a sample of 25,000 respondents aged 15 and over. The survey produces more detailed statistics on the victimization incidents experienced by the population (based on eight types of offences) and on trends in victimization. It looks at risk factors associated with victimization and rates of reporting to the police, and it evaluates the fear of crime as well as public perceptions of crime and the criminal justice system.

Methodology

The International Crime Victimization Survey is co-ordinated by the United Nations Office on Drugs and Crime (UNODC). Survey companies in different countries are selected by UNODC to conduct the survey. Data collection in Canada was conducted by Leger Marketing, a survey company in Montreal and a member of the Gallup International Association. Canada has participated in all five cycles as a result of the Research and Statistics Division in the Department of Justice having taken the lead in co-ordinating the survey.

A total of 2,000 persons aged 16 or older were selected at random from across Canada for interviewing. All interviews were conducted by telephone, using the Computer Assisted Telephone Interviewing (CATI) method. The language of the interview was either English or French. A pre-test was completed on September 22 and 23, 2004. Subsequently, all initial and follow-up interviews were completed between September 24 and October 30, 2004.

In order to ensure a high response rate and to address the potential concerns of survey respondents, a 1-800 hotline located in the Department of Justice Canada was set up. In addition, the Canadian Police Information Centre (CPIC) informed police departments across Canada that such a survey was being conducted. The survey response rate was 60%. In Canada, a total number of 4,162 telephone numbers were called for ICVS. Only 3,324 numbers were classified as "relevant or eligible contacts"⁸ Of those, 1,324 were refusals and 2,000 were completed interviews. The response rate is result of the number of completed responses by the total number of eligible contacts.

Data limitations

It is important to note that the ICVS data are estimates. They are based on information collected from a small fraction of the population. The sample sizes ranged from 1,000 to just over 7,000 respondents (Table 1). A single respondent can represent anywhere from 1,000 to over 200,000 people, depending on the country. As a result, the data are subject to sampling error. The size of the sampling error depends on many factors, including the sample size, the percentage observed, and the level of confidence chosen. The sampling error for the ICVS, using a 90% confidence interval, is shown in Table 2.

As an example of how this table should be used, consider the overall victimization rate for Canada. The ICVS estimates that 17.2% of the population was victimized in 2004. Using the table, for a sample of 2,000 and an observed percentage of 20, the sampling error is 1.5. This means that the 90% confidence interval is between 15.7% and 18.7% (17.2 ±1.5). The interpretation of the confidence interval is as follows: if repeated samples of this size were drawn, each one leading to a new confidence interval, then in 90% of the samples, the confidence interval would include the true value for the population. Because the results are subject to sampling error, the difference between some figures will not be statistically significant.

Tables

Table 1 Countries that participated in the 2004-2005 cycle of the International Crime Victim Survey and sample sizes

Table 2 Sampling error (with a 90% level of confidence)

Reference

Van Dijk, Jan, John Van Kesteren, Paul Smit, Tilburg University, UNICRI and UNODC, (2007) 'Criminal Victimisation in International Perspective: Key Findings from the 2004-2005 ICVS and EU ICS'. The Hague, Ministry of Justice, WODC.

Notes

1. International Criminal Victimization Survey and European Survey on Crime and Safety.  
2. General Social Survey (GSS) on Victimization. For specific trends and more detailed information on victimization in Canada, the GSS remains the main source of data.
3. To date, five cycles of the Survey have been conducted, in 1989, 1992, 1996, 2000 and 2004-2005, with a number of countries participating (in all, more than 60 countries have participated at least once since the first cycle). Canada has participated in all the cycles. Some countries were surveyed in 2004 and others in 2005. A large portion of the 2004-2005 data were derived from the EU ICS.
4. Van Dijk, Jan, John Van Kesteren, Paul Smit, Tilburg University, UNICRI and UNODC, (2007) 'Criminal Victimisation in International Perspective: Key Findings from the 2004-2005 ICVS and EU ICS'. The Hague, Ministry of Justice, WODC.
5. Throughout this report, any differences reported are statistically significant. When reference is made to rates or proportions being similar, the differences are not statistically significant.
6. The ICVS collects data on ten conventional crimes but also on corruption, consumer fraud including Internet and credit card fraud, drug problems and hate crimes. Also covered by the ICVS are reporting to the police, satisfaction with the police response, victim support, fear of crime, use of crime prevention measures and attitudes toward sentences.
7. Although the victimization rates are referred to as being for 2004, the incidents of victimization actually occurred in 2003. Prevalence rates for the five-year period preceding the survey are also available, but this analysis concerns only those for 2004. 
8. Numbers were not relevant for a variety of reasons including non-existent telephone numbers, calls not answered, non-response due to language barrier and non-residential numbers.