|Page tools: Print Page Print All|
Feature Article - Crime and Safety in Western Australia
The most common household offence reported in the survey was break-in, reported by 44,100 households or 6.0% of all households, while the victimisation rates for attempted break-in and motor vehicle theft were 5.1% and 1.9% respectively.
VICTIMS AND NON-VICTIMS OF CRIME, Type of Offence
The victimisation rate was highest for other households, of which 11,500 or 15.1% of such households reported at least one incident of household crime. Other households include those consisting of unrelated people, households made up of related people that do not include parents, or where another relative is residing with a family. These households account for 10.4% of all households and 13.8% of all household crime victims. Households containing a couple only recorded the lowest victimisation rate of 9.5%. Couple only households account for 24.8% of all households, and 20.6% of all household crime victims.
VICTIMS AND NON-VICTIMS OF HOUSEHOLD CRIME(a), Selected Characteristics
Of the estimated 1,459,600 residents of Western Australia aged 15 years or over at October 2000, 80,000 (5.5%) reported being victims of at least one personal crime in the 12 months to October 2000. The victimisation rate for males for personal crime was 6.2% compared with 4.8% for females. The personal crime with the highest victimisation rate was assault at 4.9%.
Victimisation rates for personal crime varied according to age and sex. Younger persons experienced higher victimisation rates than older persons, with 11.9% of persons aged 15 to 24 years experiencing personal crime. The victimisation rate progressively decreased with age to 1.3% for persons aged 55 years and over.
VICTIMS AND NON-VICTIMS OF PERSONAL CRIME(a), Sex by Age
In October 2000, there were an estimated 1,094,100 residents in the Perth Metropolitan Area accounting for 75.0% of all residents in Western Australia. Metropolitan residents accounted for 78.0% of all victims of personal crime. The victimisation rate for personal crime was 5.7% for the Perth Metropolitan Area and 4.8% for the balance of Western Australia.
VICTIMS AND NON-VICTIMS OF PERSONAL CRIME(a), Selected Characteristics
Persons in the labour force (employed and unemployed) were more likely to be victims of crime than those who were not in the labour force.The victimisation rate for unemployed persons was highest at 12.9% while the rate for employed persons was 5.7%. By comparison, the victimisation rate for those persons not in the labour force was 4.2%.
Of the 83,400 households in Western Australia that reported being victims of household crime in the twelve months to October 2000, 25,600 or 30.8%, experienced two or more occurrences. 73,600 households reported an incident of break-in or attempted break-in, and of these 28.9% reported more than one such incident.
The victimisation rate for two or more incidents of household crime was highest for attempted break-in (32.2%).
Assault was the most reported personal crime in the survey, with 35,600 persons reporting a single incident of assault and 36,300 reporting two or more incidents. 21,200 (51.7%) male assault victims reported one occurrence of assault and 19,800 (48.3%) reported two or more occurrences. Although females reported a lower overall victimisation rate for assault than males, females who reported being assaulted were more likely than males to have reported two or more incidents. Of the 30,900 females who reported assault, 53.4% experienced more than one incident.
VICTIMS OF CRIME, Number of Incidents Experienced by Type of Offence
REPORTING TO POLICE
Analysis of the most recent incident shows that the nature of the crime heavily influenced whether victims reported the occurrence to the police. For household crimes, 95.9% of motor vehicle thefts were reported to police while only 31.2% of attempted break-ins were reported. Among victims of personal crime, 27.2% of assaults were reported to police while 58.7% of robberies were reported. Among female victims, 30.8% of sexual assaults were reported. Females reported 69.6% of robberies whereas males reported 49.5%.
VICTIMS OF CRIME, Whether Police Told About Last Incident by Type of Offence
RELATIONSHIP TO OFFENDER
For occurrences of assault, the offender was known to the victim in 45,500 (or 63.3%) of cases. Where the offender was known, most incidents of assault (20.6%) involved a family member (including ex-partner) of the victim. Acquaintances (15.0%) and friends (12.5%) were the next most commonly reported known offenders.
VICTIMS OF ASSAULT, Whether Offender Known in Last Incident
Known offenders were family members or friends in almost one third (33.1%) of incidents. Where the offender was known, neighbours were involved in only 2.4% of assault incidents. In only 36.7% of incidents was the offender not known.
PERCEPTION OF CRIME/PUBLIC NUISANCE
A total of 648,900 persons (44.5%) reported no perceived crime or public nuisance problems in their neighbourhood. Of the remaining 810,700 persons (55.5%), the most commonly perceived problems were housebreaking/burglaries (40.3%), dangerous/noisy driving (30.7%) vandalism/graffiti/damage to property (29.6%), and motor vehicle theft (23.8%).
PERCEPTIONS OF CRIME OR PUBLIC NUISANCE PROBLEMS IN THE NEIGHBOURHOOD
INTERPRETING THE DATA
The terms used to describe the various types of offences in this article may not necessarily correspond with legal or police definitions. Victims are counted once only for each type of offence, regardless of the number of incidents occurring, although multiple victimisation was recorded and is included in the tables.
For the 2000 and 1999 surveys, questions were structured to enable the personal offence categories of robbery and assault to align with the National Crime and Safety Survey conducted in 1998 (see the section Comparability with other surveys for information on the National Crime and Safety Survey). Estimates for robbery and assault and consequently overall person victimisation estimates from the 1995 WA Crime and Safety Survey are not comparable with estimates from these later surveys and are not included in this article. For household crimes and sexual assault, victimisation rates from the 1995 survey are provided for comparison. For other reasons which are explained in paragraphs 12 to 13 of the Explanatory Notes, caution should be used when comparing victimisation rates between years.
1. The feature article in this publication contains results from the WA Crime and Safety Survey which was conducted throughout Western Australia in October 2000 as a supplement to the ABS Monthly Population Survey (MPS). The survey was conducted at the request of the Ministry of Justice, the WA Police Service, Ministry of the Premier and Cabinet, Office of Seniors Interests, Womens' Policy Development Office, and the Crime Research Centre, University of Western Australia.
2. Information was collected from individuals and households about their experience of selected crimes, reporting behaviour and individuals’ perception of crime problems in their neighbourhood.
3. The MPS was conducted during the two weeks commencing Monday, 9 October 2000.
4. Information was sought from a 7/8 sample of the MPS which was approximately 6,800 persons, of whom 5,541 (81%) responded. Approximately 3,470 households were surveyed and complete household data were obtained from 2,787 (80%) of these.
5. Estimation and imputation procedures were employed to reduce the effect of non-response.
6. The survey was conducted for all persons aged 15 and over who were usual residents of private dwellings, except:
7. Residents of other non-private dwellings such as hospitals, motels and prisons were excluded from this survey.
8. Coverage rules were applied to ensure that each person was associated with only one dwelling, and hence had only one chance of selection in the survey.
9. Crime and Safety questionnaires were either delivered to the selected households by ABS interviewers or, in the majority of cases, mailed out to households who were interviewed for the MPS over the telephone. The questionnaires were completed by household members and returned to the ABS by mail. Each household received:
COMPARISONS WITH POLICE STATISTICS
10. Responses obtained in this survey are based on the respondent's perception of their having been the victim of an offence. The terms used summarise the wording of the questions asked of the respondent, and may not necessarily correspond with the legal or police definitions which are used for each offence.
CHANGES TO THE QUESTIONNAIRE
11. For the 2000 WA Crime and Safety Survey, questions about robbery and assault incidents were included to be consistent with the 1999 WA Crime and Safety Survey and with the 1998 National Crime and Safety Survey (see the section Comparability with other surveys for information on the National Crime and Safety Surveys). The data for robbery and assault are therefore not comparable with data collected in the 1995 WA Crime and Safety Survey. The definitions of robbery and assault for the statistics included in this publication can be found in the Glossary.
Caution when comparing 1995 and either 1999 or 2000 victimisation rates
12. Caution should be exercised when comparing data and victimisation rates between 1995 and either 1998, 1999 or 2000. Significant changes to the question wording for incidents of both robbery and assault have made data comparisons for personal crime invalid, these are further detailed in the section Comparability with other surveys.
13. While the standard error formulae provided in the section Standard errors may be used to calculate the standard error on the difference between estimates for two different years, the survey is not specifically designed to measure this type of movement. To do so would require a high proportion of common persons selected in the sample for each survey year. Comparisons of this nature should therefore be made with caution. When estimating the standard error of a movement between years, the movement standard error will be approximately 1.4 times the standard error on the level estimate, if the standard errors on the two level estimates are similar.
14. The ABS produces a wide range of publications containing social and demographic statistics. Other ABS publications which relate to this survey topic are shown below.
COMPARABILITY WITH OTHER SURVEYS
Some of the more important methodological, definitional and other differences between the 1995, 1998, 1999 and 2000 WA Crime and Safety Surveys follow.
All the 1995 questionnaires were delivered to the respondent households by interviewers at the time of their Labour Force Survey (LFS) interview. In 1999 and 2000, most of the questionnaires were mailed out from the Perth Office of the Australian Bureau of Statistics within two weeks of the completion of LFS interviewing. Similarly, the 1998 questionnaires were mailed out within two weeks of LFS interviewing. There were few differences in methodology between the 1999 and 2000 surveys.
In the 1998, 1999 and 2000 surveys, use was made of the ‘please describe’ responses to refine the coding of all crimes. This was not possible in 1995. As a result, coding of crimes may be less accurate in 1995 compared with other years.
Motor vehicle theft in 1998, 1999 and 2000 includes all motor vehicles, whereas the 1995 survey included only registered motor vehicles.
For 1995 information on the personal crimes of robbery and assault was collected using a different set of questions.
Due to the significant changes in question wording, it is not possible to provide comparable figures on the personal crimes of robbery and assault, therefore the 1995 data have not been included in this publication.
The robbery and assault questions used in 1998, 1999 and 2000 were aimed at obtaining more detail on what actually happened to the victims, so as to give a better picture of the nature and seriousness of the incident, and to allow easier comparison with data from other sources. As a result of the additional information collected in these years, incidents were able to be more accurately recorded as a robbery or an assault.
It is believed that the 1995 robbery figure included a number of incidents that police would classify as theft and some others that would be classified as assault. As a result of the problems with this question, questions were added in 1998, 1999 and 2000 to determine whether or not anything was stolen and to determine what actually happened in the incident.
In 1998, 1999 and 2000, a two-phase approach was used, as indicated in the above table. Note that the 1999 and 2000 robbery figures include attempts, which are specifically excluded from the 1995 WA Crime and Safety Survey.
A number of differences exist between the 1995, 1998, 1999 and 2000 questions, in particular the prompt in 1998, 1999 and 2000 to include assaults from people the respondent knew well. Note also that the 1998, 1999 and 2000 assault question specifically includes attempts whereas the 1995 survey question did not. However, it is believed that the 1995 data are likely to include some attempts. Where an incident of assault has been reported in the 1998, 1999 and 2000 surveys, there is a data item which specifies whether the incident involved the use, attempted use or threat of force or violence. This information was used to refine the assault data.
In 1998, 1999 and 2000, a two-phase approach was also used for assault, as indicated in the above table. Note that the 1998, 1999 and 2000 assault figures include attempts, which are specifically excluded from the 1995 WA Crime and Safety Survey.
REASONS FOR NOT TELLING THE POLICE ABOUT AN INCIDENT
In 1995, this question was open-ended and the information given was coded to a set of responses which reflected those most frequently recorded. In 1999 and 2000, a tick box response list was used which was developed from the responses obtained to this question over a number of surveys, and room was provided for other reasons to be written down. It is possible that a respondent’s interpretation of the precoded responses may be different to how an ABS coder would have coded a written response from the respondent. Tick boxes were specifically used in an attempt to reduce the number of uncodable responses. There were proportionally less reasons coded to ‘other’ in 1999 and 2000 compared with 1995.
1. The WA Crime and Safety Survey provides information on the incidence of selected categories of crime and crime reporting behaviour for persons aged 15 and over for the 12 months to October 2000. In addition, females aged 18 and over were asked to provide information on personal experiences of sexual assault.
2. A customised data service is available to meet special data requirements.
3. Information collected in the survey includes:
Types of offences:
Perceived crime or public nuisance problems.
4. Data can be classified by the following variables:
5. To discuss your data requirements, or for further information regarding this survey, please contact the Statistical Consultancy Unit on Perth (08) 9360 5947.
1. For this survey, the effects of non-response were investigated by analysing the demographic composition of the Monthly Population Survey sample. This information was used to determine the appropriate adjustment procedure for non-response. An initial person weight that accounted for the probability of selection and non-response was then formed.
2. The harmonic mean of the initial weights of the persons that reside in the household was then used as the initial household weight. This initial household weight was then calibrated against independent estimates of population (benchmarks) for persons and households to obtain a common weight. In this survey, broad age by sex benchmarks were used for persons and part of state for households.
3. Expansion factors, or weights, are values by which information for the sample is multiplied to produce estimates for the whole population. From this survey, estimates are produced referring to persons, and to households, and the weights are calculated so that each person in a household has the same weight and that weight is also used for the household.
4. Estimates of counts are then simply obtained by summing the weights of either households or persons within the required group. For example, an estimate of the total persons who were robbed in the population would be obtained by simply adding the weight for each person that was robbed in the sample.
RELIABILITY OF ESTIMATES
5. Estimates in this publication are subject to non-sampling and sampling errors.
6. Non-sampling errors may arise as a result of errors in the reporting,recording or processing of the data. Non-sampling errors can be introduced through inadequacies in the questionnaire, non-response, inaccurate reporting by respondents, errors in the application of survey procedures, incorrect recording of answers and errors in data entry and processing.
7. It is difficult to measure the size of the non-sampling errors. The extent of these errors could vary considerably from survey to survey and from question to question. Every effort is made in the design of the survey and development of survey procedures to minimise the effect of these errors.
8. Sampling error is the error which occurs by chance because the data were obtained from a sample, rather than the entire population.
ESTIMATES OF SAMPLING ERROR
1. One measure of the variability of estimates which occurs as a result of surveying only a sample of the population is the standard error, shown in the table Standard errors of estimates of households and persons - October 2000.
2. There are about two chances in three (67%) that a survey estimate is within one standard error of the figure that would have been obtained if all households/persons had been included in the survey.There are about nineteen chances in twenty (95%) that the estimate will lie within two standard errors.
3. Linear interpolation is used to calculate the standard error of estimates falling between the sizes of estimates listed in the table.
4. The standard error can also be expressed as a percentage of the estimate. This is known as the relative standard error (RSE). The RSE is determined by dividing the standard error of an estimate SE(x) by the estimate x and expressing it as a percentage. That is -
(where x is the estimate). The relative standard error is a measure of the error (relative to the size of the estimate) likely to have occurred due to sampling.
5. Proportions and percentages formed from the ratio of two estimates are also subject to sampling error. The size of the error depends on the accuracy of both the numerator and the denominator. The formula for the relative standard error of a proportion or percentage is -
6. For all tables in this publication, only estimates with relative standard errors of 25% or less, and percentages based on such estimates, are considered sufficiently reliable for most purposes. Estimates and percentages with relative standard errors between 25% and 50% have been included, preceded by the symbol * to indicate that they are subject to high standard errors and should be used with caution.
STANDARD ERRORS OF ESTIMATES OF HOUSEHOLDS AND PERSONS - OCTOBER 2000
PRINTER FRIENDLY VERSION OF "CRIME AND SAFETY IN WESTERN AUSTRALIA"
These documents will be presented in a new window.