Kesteren, J.N.van, Mayhew, P. &
Nieuwbeerta, P. (2000)
Criminal Victimisation in Seventeen Industrialised Countries:
Key-findings from the 2000 international Crime Victims Survey.
the Hague, Ministry of Justice, WODC. Onderzoek en
beleid, nr. 187
download text in pdf
The International Crime Victimisation Survey (ICVS) is the most far-reaching programme of fully standardised sample surveys looking at householders' experience of crime in different countries. The first ICVS took place in 1989, the second in 1992, the third in 1996 and the fourth in 2000. Surveys have been carried out in 24 industrialised countries since 1989, and in 46 cities in developing countries and countries in transition. This report deals with seventeen industrialised countries which took part in the 2000 ICVS.
The reason for setting up the ICVS was the inadequacy of other measures of crime across country. Figures of offences recorded by the police are problematic due to differences in the way the police define, record and count crime. And since victims report most crimes the police know about, police figures can differ simply because of differences in reporting behaviour. It is also difficult to make comparisons of independently organised crime surveys, as these differ in design and coverage.
For the countries covered in this report, interviews were mainly conducted by telephone (with samples selected through variants of random digit dialling). The overall response rate in the 17 countries was 64%. Samples were usually of 2,000 people, which mean there is a fairly wide sampling error on the ICVS estimates. The surveys cannot, then, give precise estimates of crime in different countries. But they are a unique source of information and give good comparative information.
Each participating country paid for its own fieldwork. The Dutch Ministry of Justice also provided financial assistance for overheads.
Technical aspects of the surveys in many countries were co-ordinated by a Dutch company, Interview-NSS, who sub-contracted fieldwork to local survey companies. The NSCR and Leiden University managed survey results. The results in this report relate mainly to respondents' experience of crime in 1999, the year prior to the 2000 survey. Those interviewed were asked about crimes they had experienced, whether or not reported to the police. The main results follow.
The ICVS allows an overall measure of victimisation which is the percentage of people victimised once or more in the previous year by any of the eleven crimes covered by the survey. This prevalence measure is a simple but robust indicator of overall proneness to crime. The countries fall into three bands.
- Above 24% (victim of any crime in 1999): Australia, England and Wales, the Netherlands and Sweden
- 20%-24%: Canada, Scotland, Denmark, Poland, Belgium, France, and USA
- Under 20%: Finland, Catalonia (Spain), Switzerland, Portugal, Japan and Northern Ireland.
For countries in previous sweeps of the ICVS, the present results generally mirror previous ones as regards relative rankings. In terms of the number of crimes experienced per 100 people (an incidence rate), the picture is slightly different. The USA fares relatively worse on incidence rates than on prevalence rates. In contrast, the position of Denmark and Canada slip down somewhat. Incidence rates are highest in England and Wales, Australia and the Netherlands.
The risk of having a car stolen was highest in England and Wales (2.6% of owners had a theft), Australia (2.1%), and France (1.9%). Japan, Switzerland, Catalonia, the USA, Finland, and the Netherlands show risks of 0.5% or less. Those in Poland, Japan, Belgium and the Netherlands were least likely to get their cars back - indicating proportionately more professional theft. Recovery rates were above 80% in Sweden, Australia, and the USA - indicating more thefts for 'joyriding'. In the
eleven countries with surveys in 1996 and 2000, there is little change in the proportion of stolen cars recovered, but it is now lower than it was in 1992, probably indicating a general move towards more professional theft since then.
Having something stolen from a car (e.g., luggage, radios, car mirrors etc.) was more common. Highest risks were in Poland (9% of owners had one or more theft), England and Wales (8%), Australia (7%) and the USA (7%). Risks were lowest in Denmark, Finland, the Netherlands, Switzerland and Japan: 4% or less were victimised. Car vandalism was most common in Scotland (12% of owners had their car damaged), Poland, England and Wales, the Netherlands and Australia (about 10%-11%). Risks were low in Northern Ireland, Japan, Denmark and Switzerland (less than 5%). Countries with higher rates of car vandalism generally had higher rates of thefts from cars, but the association between vandalism and thefts of cars was weaker.
Highest risks of motorcycle and moped theft were in Denmark and England and Wales (4% of owners were victimised). Although Japan has low risks for most crimes, thefts of motorcycles was comparatively high (3%).
For bicycle theft, the highest risks were in Japan, Sweden, the Netherlands and Denmark (about 8% of owners had a bicycle taken). The 2000 ICVS results suggest - as in previous years - that bicycle theft is highest in countries where there are most bicycles owned: ie, a plentiful supply of targets encourages rather than dampens theft demand. Also, where bicycle theft is common, stealing cars occurs less often - though the relationship is rather less strong than in earlier ICVS sweeps.
The proportion of households who had a completed or attempted burglary was highest in Australia (7%), England and Wales (5%), Canada, Denmark and Belgium (all 4%). The pattern of relative risk is reasonably similar whether the focus is on burglary with entry or attempts. Where burglars are
successful in gaining entry, they are also more active in trying to do so. Nonetheless, the proportion of burglaries that involved attempts varied somewhat by country. More attempts at entry failed in Finland, France, Belgium, Scotland, the Netherlands, England and Wales, Switzerland and the USA (all had above 50% attempts).
With the exception of Finland, the ICVS evidence suggests that homes in these countries are better protected by security devices. This may explain why burglars more often fail to gain entry.
Theft of personal property
Thefts of personal property will be heterogeneous in nature, but the highest risks were in Australia, Sweden, and Poland (about 5%-6% of people were victimised).
In roughly a third of thefts, the victim was carrying what was stolen - termed 'pickpocketing'. Risks of pickpocketing were most common in Poland (4%). Risks were also comparatively high in Belgium, the Netherlands, Denmark, Catalonia, and England and Wales (about 2%). As previous sweeps have found, risks were lowest outside Europe: in Japan, Canada, and the USA.
An overall measure of contact crime was taken as robbery, assaults with force, and sexual assaults (against women only). The highest risks were in Australia, England and Wales, Canada, Scotland and Finland: over 3% were victims. This was more than double the level in USA, Belgium, Catalonia, Portugal, and Japan (all under 2%). In Japan the risk of contact crime was especially low (0.4%).
Robbery was comparatively uncommon in all countries. Risks were highest in 1999 in Poland (1.8%), England and Wales, and Australia (both 1.2%). By far the lowest risks were in Japan and Northern Ireland (0.1%). On average, just over a third of victims of robbery said the offender(s) carried a weapon of some sort - in most cases a knife. There was a higher than average use of weapons in the USA, Catalonia, Scotland, and Portugal. Although not very statistically robust, the data indicate that guns were used relatively more often in Catalonia and the USA.
Two types of sexual incidents were measured: offensive sexual behaviour and sexual assault (i.e. incidents described as rape, attempted rape or indecent assaults). For all countries combined, just over one per cent of women reported offensive sexual behaviour. The level was half that for sexual assaults. Women in Sweden, Finland, Australia and England and Wales were most at risk of sexual assault. Women in Japan, Northern Ireland, Poland and Portugal were least at risk. Many of the differences in sexual assault risks across country were small. Generally, the relative level of sexual assault in different countries accorded with relative levels of offensive sexual behaviour - though there were a few differences.
Women know the offender(s) in about half of the all sexual incidents: in a third they were known by name, and in about a sixth by sight. (More assaults involved offenders known by name than did incidents of offensive sexual behaviour.) Most sexual incidents involved only one offender. Weapons were very rarely involved.
Assaults and threats
Taking all countries together, 3.5% were victims once or more of assaults or threats in 1999. Risks were highest in Australia, Scotland, England and Wales (about 6%) and Canada (5%). Risks were lowest in Japan, Portugal, (under 1%) and Catalonia (1.5%). Offenders were known in about half the incidents overall. Men were less likely to know offenders than women. Weapons (especially knifes) were said to have been used (if only as a threat) in just under a quarter of incidents.
Country profiles of crime
Taking all countries together, car vandalism forms a full quarter of crimes experienced by ICVS respondents. Car vandalism - together with thefts of and from cars - means that over 40% of ICVS crimes involve cars. Contact crimes comprise about a quarter of the crimes measured, most of them assault and threats. Motorcycle and bicycle theft, burglaries, and thefts of personal property each contributed just over 10% overall. The largest difference between countries was with regard to the bicycle theft 'share', reflecting varying ownership rates.
The make-up of crime differs across country. Catalonia and Portugal stand out against the norm in having a crime problem dominated by incidents involving cars: rather more than 60% of all the crimes counted. Japan was also unusual in that 40% of the crime counted by the ICVS involved thefts of two-wheelers. The distinctive feature of Finland was the unusually high share of all crime that sexual incidents accounted for (over a quarter).
Victims were asked to assess the seriousness of what happened. Mean seriousness scores were computed for different offence types. Car thefts where the car was not recovered were viewed most seriously. Next most serious were sexual assaults, then car thefts even if the car was recovered, and robbery involving a weapon. Assaults with force were scored much on a par with burglaries with entry. The least serious crimes were car vandalism, theft from cars, and bicycle theft. Results in previous ICVS sweeps were similar.
Overall mean score did not differ much by country. This suggests that people in different countries have similar attitudinal thresholds about the seriousness of different crimes. It also suggests people do not differ very greatly in the types of incidents they tell interviewers about. The ranking of offences in seriousness terms were also very similar, again indicating a high degree of consensus about the import of conventional crimes.
We corrected the victimisation rates for crime seriousness to see how countries fared on a crime count taking seriousness into account. It did not greatly alter the 'burden of crime' picture from other measures. Australia, England and Wales, the Netherlands and Sweden still remain most pressured by crime. However, Denmark and Canada fall back in the relative order when seriousness is taken into account, while the USA and Northern Ireland go higher up the list.
Trends in crime
Generally speaking, the ICVS suggests that crime rose between 1988 and 1991, stabilised of fell in 1995, then fell back more in 1999. This is the dominant pattern in many individual countries.
The picture in North America differs from that in Europe. Crime levels are lower than in 1988. In the three European countries with four ICVS measures (England and Wales, Finland, and the Netherlands), crime levels are still higher than in 1988. Compared to 1991, risks also fell more in North America than in five of the seven European countries showing falls. Since 1995, there has been more consistent falls in property crime. Changes in violent crime are variable.
Differences in risks of crime
Risks for different social groups were examined using multivariate analysis. All 17 countries were taken together. Those in the largest conurbations (of populations over 100,000) were most at risk. Net of other effects, risks were 60% higher in the most urban areas compared to the least urbanised ones. The biggest differentials were for sexual incidents and thefts of and from cars ('car thefts').
Households with higher incomes were more at risk those poorer ones - by a third. The biggest differential was for car thefts. The difference for burglary was lowest (higher income households were about 10% more at risk). The analysis here is focused on individual risks rather than area ones. In poorer neighbourhoods, households in general might have higher risk, but more affluent households nonetheless emerge as the most vulnerable. They may offer more 'criminal 'rewards'.
Younger respondents were more at risk than older ones. Risks of all ICVS crimes were well over double than for those aged 55 or more.
Those who went out more frequently were rather more vulnerable - by about 20%. Those who were unmarried were also more at risk, net of other effects. Risks of contact crime were double than those for people in permanent relationships. For robbery and assaults and threats men were about 20% more at risk than women were.
Reporting to the police
Nearly all cars and motorcycles stolen were reported, as were burglaries with entry. About two-thirds of thefts from cars and bicycle thefts were reported, but on average only nearly half of attempted burglaries and robberies were. Reporting rates for other crime types were lower.
We took six crime types to look at differences in reporting levels. The highest reporting rates were in Denmark and Sweden, Northern Ireland, the Netherlands (60% or just under). Reporting was also relatively common in Belgium, England and Wales, Switzerland, France and Scotland (above 50%). Crimes were least often reported in Portugal, Japan, Catalonia, and Poland (less than 40%).
The most common reason for not reporting in all countries was that the incident was 'not serious enough' or there was 'no loss'. (Five crimes were considered.) The idea that the police could do nothing about what happened also featured fairly frequently. Few victims mentioned fear or dislike of the police as a reason for reporting, though it was more common in relation to contact crime. Victims were also asked why they did report. Victims of sexual incidents and assaults and threats were most concerned to stop what happened being repeated. For burglaries with entry and thefts from cars, more than a third reported because they wanted help in getting property back, and a third did so for insurance reasons. Four in ten victims referred to the civic obligation to notify the police.
The majority of victims were satisfied with how the police responded to their crime report. Highest levels of satisfaction were in Denmark, Catalonia and Switzerland. The police response was considered least good in Portugal, Poland, France and Japan. The main reason for being unhappy with the police response was that they 'they did not do enough': half complained about this. About a third felt that the police 'were not interested'. Some victims were asked whether they got help from a specialised victim support agency. Support was more often given to victims of contact crimes (10% were offered help) than to victims of burglary (5%). Victims in the UK were offered most support. There were also comparatively high level of support in the Netherlands, Canada, Sweden, the USA, and Denmark. Least support seems to have been available to victims in Portugal, Japan, Finland, France and Poland. In most countries, around one in three burglary victims who had not been given help would have welcomed some. Four in ten victims of contact crime felt this too.
People were asked to say whether or not the police did a good job in controlling crime in their area, and whether the police were helpful. Police performance was most favourably judged in the USA and Canada. Satisfaction levels were also comparatively high in Scotland and Australia. The poorest judgements were from this in Portugal, Poland the Netherlands, Japan and Catalonia. In most countries, police performance was judged less favourably after 1988. But compared to the 1996 ICVS, police performance was rated more highly in 2000 in all but one (Sweden) of 11 countries with measures for both years.
Anxiety about burglary
On average, nearly a third of people felt they were likely or very likely to be burgled in the coming year. Those in Portugal (58%), Belgium, and France (about 45%) were most pessimistic. There was least concern in the Scandinavian countries, the USA, and the Netherlands (under 20% thought a burglary was likely).
Concern about burglary rose between the 1989 and 1992 ICVS, and has since fallen - although a few countries are exceptions. Falling perceptions of the likelihood of burglary broadly match trends in ICVS burglary levels.
Safety on the streets
When asked how safe they feel walking alone in their area after dark, on average just under a quarter felt very or a bit unsafe. Those in Catalonia, Australia and Poland were most anxious (about a third felt a bit or very unsafe). Next highest levels were in Portugal and England and Wales. Feelings of vulnerability were lowest in the USA and Sweden, although there were several other countries with only marginally higher figures.
Whereas anxiety about burglary to some extent matches national risks, feelings about street safety are not consistently related to levels of 'street trouble'. The lack of much relationship between anxiety and risks of street crime has been evident in previous ICVS results. It may mean that fear of street crime is determined by specific 'cultural' pressures.
Safety at home
A new question in the 2000 ICVS asked about safety at home. A much smaller proportion felt unsafe at home (6% overall felt a bit or very unsafe). Those in Poland felt most insecure (15%), followed by Portugal, Japan and Belgium.
Summary measures of home security were taken as the proportion of homes with burglar alarms and special door locks. The highest alarm ownership was in England and Wales, Australia, Scotland, the USA, Canada, and Belgium. Generally, countries with highest alarm ownership also had more homes with special door locks The use of preventive measures is increasing in most countries.
Attitudes to punishment
People were asked about the most appropriate sentence for a recidivist burglar aged 21. A community service order was seen the most appropriate sentence overall: 41% recommended it. But there was a wide spread of opinion. It was the first choice in half of the 16 countries considered, with support strongest in France, Catalonia (two-thirds recommending it), Belgium, Poland and Portugal (over half). Fewer than 30% opted for a community sentence in the UK, and fewer still in Japan and the USA. Imprisonment was recommended by 34% of respondents overall, and was the first choice in eight countries.
Support for imprisonment in 2000 was strongest in the USA, the UK, and Japan (all with about half or more choosing it). Those in Catalonia and France were least in favour of imprisonment.
The 2000 ICVS shows a general hardening of attitudes towards punishment, with increasing proportions supporting imprisonment. The most marked changes have been in Canada, England and Wales, Scotland, the Netherlands, and Sweden.
download full text in pdf
Kesteren, J.N.van, Mayhew, P. & Nieuwbeerta, P. (2000)
Criminal Victimisation in Seventeen Industrialised Countries:
Key-findings from the 2000 international Crime Victims Survey.
the Hague, Ministry of Justice, WODC. Onderzoek en beleid, nr. 187