Recorded Crime - Victims, Australia methodology

1 This publication presents national statistics relating to victims of crime for a selected range of offences as recorded by police. These offences may have been reported by a victim, witness or other person, or they may have been detected by police.

2 The statistics presented in this publication are not designed to provide a total count of unique victims, nor the total number of individual offences that came to the attention of police. For further information on the counting methodology refer to paragraphs 14–20.

3 Statistics in this publication are derived from information held in administrative systems which are collected and maintained by police agencies within each state and territory. This information is collected by the ABS and compiled according to the National Crime Recording Standard (NCRS) in order to maximise consistency between states and territories. For information on the NCRS refer to paragraphs 66–73.

4 National crime statistics are produced annually on a calendar year basis. The reference period for this publication relates to offences that have been reported to police between 1 January and 31 December 2018. Additionally, selected statistics are available for the 2010 to 2018 reference periods.

5 The scope of this collection includes victims of offences classified to selected divisions and/or subdivisions of the Australian and New Zealand Standard Offence Classification (ANZSOC). A victim can be a person, premises, organisation, or motor vehicle depending on the type of offence (for more information refer to paragraphs 14–20). Selected offence categories are:

• Homicide and related offences (including Murder, Attempted murder and Manslaughter, but excluding Driving causing death and Conspiracy to murder)
• Assault
• Sexual assault
• Abduction and kidnapping
• Robbery
• Blackmail and extortion
• Unlawful entry with intent/burglary, break and enter
• Motor vehicle theft
• Other theft

6 The above offence categories generally correspond to divisions/subdivisions of the ANZSOC:

• Homicide and related offences is ANZSOC division 01
• Assault is ANZSOC subdivision 021
• Sexual assault is ANZSOC subdivision 031
• Abduction and kidnapping is ANZSOC subdivision 051
• Robbery is ANZSOC subdivision 061
• Blackmail and extortion is ANZSOC subdivision 062
• Unlawful entry with intent/burglary, break and enter is ANZSOC division 07
• Motor vehicle theft is made up of ANZSOC groups 0811 and 0812
• Other theft is made up of ANZSOC groups 0813, 0821, 0823, 0829 and 0841

7 For ease of reading, some offence categories have been abbreviated for use in both commentary and data cubes as follows:

• Abduction and kidnapping appears as ‘Kidnapping/abduction’
• Blackmail and extortion appears as ‘Blackmail/Extortion’
• Unlawful entry with intent/burglary, break and enter appears as ‘Unlawful entry with intent’ or ‘UEWI’

8 Statistics in this publication relate to both completed and attempted offences, i.e. those where the intent is not fulfilled. Attempts to commit an offence are classified to the same ANZSOC divisions and/or subdivisions as completed offences, with the exception of:

• Murder, where murder and attempted murder are distinguished as separate offence categories
• Attempted motor vehicle thefts due to difficulties in distinguishing these offences from criminal damage
   

9 The scope excludes the following:

• Conspiracy offences
• Threats to commit an offence (An exception to this exclusion is assault where there is an apprehension that the direct threat of force, injury, or violence could be enacted, which is in-scope of the collection. This also applies to offences like robbery, kidnapping/abduction and blackmail/extortion where an element of threat is implicit in the nature of the crime)
• Aid, abet and accessory offences
• Deprivation of liberty offences

10 Where an outcome of investigation determines ‘no crime’ was committed i.e. the offence was reported to police but later deemed to be unfounded, false, or baseless; counts are excluded from the data. For further information see paragraphs 27–28.

11 Victims of crime as recorded by the Australian Federal Police (AFP) are out of scope of this publication. As such, victims of crime in Australia’s ‘other territories’ such as Christmas Island, the Cocos (Keeling) Islands and Jervis Bay Territory are out of scope as these territories are under AFP jurisdiction.

12 The offence categories used within this publication are based on the Australian and New Zealand Standard Offence Classification (ANZSOC), (third edition) (cat. no. 1234.0). ANZSOC provides a uniform national framework for classifying offences across Australia and New Zealand for statistical purposes. The classification is a hierarchical structure allowing for varying degrees of detail to be published depending on the level of detail in the source information. The classification structure is provided as an Appendix.

13 This publication uses a number of other classifications to categorise variables. The classification structures for location, weapon use, relationship of offender to victim and outcome of investigation have been provided as an Appendix.

14 A victim can be a person, premises, organisation or motor vehicle depending on the type of offence. A victim of a criminal incident is classified to one of the offence categories in scope of this collection (see paragraphs 5–11 for offences in scope).

15 Data are compiled on the basis of the date an offence is reported to police and recorded within the reference period. This corresponds to either the date the offence was reported to police by a member of the public or when it was detected by police, and was recorded on police systems. The report date may not necessarily be the date when the offence occurred. This is particularly the case for Homicide and related offences and Sexual assault, where in some instances there may be a large time difference between when the offence(s) occurred and the report/detection date.

16 It should be noted that the Recorded Crime – Victims collection does not enumerate unique persons or organisations. If a person is a victim of one offence, they will be counted once in this publication. If a person:

• is a victim of the same offence multiple times in the same incident, the victim is counted once.
• is a victim of the same offence multiple times in different incidents throughout the year, the victim is counted once for each incident.
• is a victim of multiple offences in the same incident that fall within the same offence category, the victim is counted once, and the lowest ANZSOC code recorded within that offence category is recorded as the offence.
• is a victim of multiple offences that fall in different offence categories, the victim is counted once in each of the different categories, meaning one victim can be presented multiple times under different offence categories.

17 The following diagram illustrates the counting rules for the Recorded Crime – Victims collection. It demonstrates how one unique person who has been a victim of five offences is counted three times, due to the types of offences committed.

18 Other examples of how different victim scenarios are counted are explained below:

• If a victim is assaulted by several offenders or a victim is repeatedly assaulted by the same offender, but reports the victimisation to police as part of the same incident, the victim would be counted once for assault.
• On the other hand, if a victim reports these offences to police as separate incidents, then a count is made for each separate report.
• If a bank with several customers present is robbed, one robbery is counted, with the victim being the bank. If personal property is also taken from two customers, there would be three victims; the bank and the two customers.
• One victim is counted for each motor vehicle stolen. For example, if five cars are stolen from a car yard, this is counted as five motor vehicle thefts.

19 For the offence of Unlawful entry with intent (UEWI) the following applies:

• One victim is counted for each place/premises victimised, which can consist of either a single structure (e.g. house), part of a single structure (e.g. flat), or multiple structures (e.g. farmstead with house, barns and sheds). The same property containing the same structure(s) can be counted differently depending on the occupancy arrangements at the time.
• For multiple structures on the same property with the same occupant(s), one victim is counted regardless of the number of separate structures unlawfully entered with intent. This would apply to a house with attached or unattached garage and a backyard shed located on the one property; and warehouses occupied by a sole organisation located on the same property.
• For multiple structures on the same property, but occupied by more than one household or organisation, one victim is counted for each separate household or organisation. Where a business premises has an attached residence that is occupied by the same person(s), the registered business is considered to be a separate victim.
• In the case of UEWI to individual areas in a building that is rented, leased or occupied separately, one victim is counted for each separate tenant/owner. For example, in a block of 10 flats leased by 10 different tenants where three flats are unlawfully entered, there would be a count of three UEWIs. If unlawful entry to the building itself is recorded, an additional offence of UEWI to that building is counted. This instance would apply to apartments in one building, offices of several commercial firms in one business building, shops in a shopping complex, hotel rooms and lodging houses.

20 The definition of victim varies between offence types, see Glossary for further information.

21 The age of a victim as presented in this publication refers to the age the victim was at the time they became known to police, rather than the age that the person was when they experienced victimisation. For example, if a victim was sexually assaulted at age 14 years but did not report the offence to police until they were 18 years of age, their age in this publication would be 18 years.

22 Preliminary analysis indicates that the proportion of historical offences present in the Recorded Crime – Victims data differs by offence grouping. Sexual assault counts in particular, have the highest proportion of offences that were committed more than 12 months ago. Care should be exercised when interpreting age data for Sexual assault victims.

23 Indigenous status is based on self-identification by the individual who comes into contact with police. The quality of the data are dependent on police asking individuals to self-identify and responses being recorded on police systems. Where individuals are not able to provide an answer for themselves, jurisdictions may accept a response where a next of kin/guardian provides the information. Where neither occurs, the victim is recorded with an Indigenous status of ‘Not stated’.

24 This publication presents Indigenous status data for a selected range of personal offences for New South Wales, Queensland, South Australia and the Northern Territory. Based on an ABS data quality assessment, Indigenous status data for other jurisdictions are not of sufficient quality for national reporting within the Recorded Crime -Victims collection. 

25 The proportion of victims with an Indigenous status of ‘Not stated’ varies by offence type

26 For information about Aboriginal and Torres Strait Islander victimisation rates refer to paragraphs 36–39.

27 Data are presented in this publication for outcome of investigation at 30 days after the date of report. However, where police have determined after an investigation that 'no crime' has occurred at 30, 90 or 180 days since the initial report to police, the offence is excluded from the data.

28 The Northern Territory is unable to code their outcome of investigation data in line with the requirements of the national outcome code 'no crime'. As a result, Northern Territory data may include victim counts for those situations where police have determined after investigation that 'no crime' has occurred.

29 The relationship of offender to victim information is initially recorded as the relationship as perceived by the victim at the time of the offence, with some jurisdictions updating this data item as the investigation progresses.

30 Relationship of offender to victim data for Western Australia are not of sufficient quality for national reporting.

31 There are some inconsistencies in relationship of offender to victim data across jurisdictions:

• New South Wales is unable to provide relationship data for robbery offences.
• Victoria records the relationship of the victim to offender rather than the offender to the victim, and data are subsequently re-coded to meet the requirements of the Recorded Crime – Victims relationship classification.

32 There are some inconsistencies in the coding of current and former boyfriends and girlfriends across the jurisdictions, which should be taken into account when making comparisons:

• Boyfriend/girlfriend:
  • For New South Wales, Victoria, Queensland, South Australia and Tasmania, boyfriend and girlfriend are coded to ‘boyfriend/girlfriend’
  • For the Northern Territory, some boyfriends and girlfriends may be included in 'Other non-family member n.e.c.' or in 'Partner'
  • For the Australian Capital Territory, boyfriend and girlfriend are coded to 'Partner'
• Ex-boyfriend/ex-girlfriend:
  • For Victoria, Queensland, South Australia and Tasmania ex-boyfriends and ex-girlfriends are coded to ‘ex-boyfriend/ex-girlfriend’
  • For New South Wales, ex-boyfriends and ex-girlfriends are coded to 'Boyfriend/girlfriend'
  • For the Northern Territory and the Australian Capital Territory, ex-boyfriends and ex-girlfriends are coded to ‘Ex-partner’

This has the following impacts on data as presented in the publication tables:

• Tables containing the categories of ‘Family member’ and ‘Non-family member’:
• New South Wales data for family member may be overstated and data for non-family member understated as ex-boyfriend and ex-girlfriend are included in 'Boyfriend/girlfriend'.
• Northern Territory data for family member may be understated and data for non-family member overstated as some boyfriends/girlfriends may be included in 'Other non-family member n.e.c.'
• Tables containing the category ‘Intimate partner’: Northern Territory data for the category may be understated as some boyfriends/girlfriends may be included in 'Other non-family member n.e.c.'
   

33 Victimisation rates are expressed as victims per 100,000 of the ABS Estimated Resident Population (ERP). They are calculated using the ERP as at the midpoint of the reference period (i.e. 30 June).These rates generally accord with international and state and territory practice.

34 The rates presented in this issue are calculated using ERP data based on the 2016 Census of Population and Housing. For population estimates and information on the methodology used to produce ERP, see Australian Demographic Statistics (cat. no. 3101.0). For Recorded Crime – Victims data, all estimates and projections for the Australian Capital Territory exclude Jervis Bay Territory. All estimates and projections for Australia exclude the external territories of Christmas Island and the Cocos (Keeling) Islands.

35 Victimisation rates vary across sex and age categories. This publication presents age and sex specific victimisation rates, which are calculated using estimates of the age and sex breakdowns of the population.

36 Victimisation rates for the Aboriginal and Torres Strait Islander and non-Indigenous persons are calculated using recast estimates (for the period 1996 to 2011) and projections (2012 to 2026) of the Aboriginal and Torres Strait Islander population based on data from the 2011 Census of Population and Housing, and rely upon assumptions about future fertility, paternity, life expectancy at birth and migration.

37 The Aboriginal and Torres Strait Islander projections used are based on Series B in the ABS publication Estimates and Projections, Aboriginal and Torres Strait Islander Australians, 2001 to 2026 (cat. no. 3238.0).

38 Rates for the non-Indigenous population are calculated using the ERP for the total population of the state or territory minus the projected Aboriginal and Torres Strait Islander population for the relevant jurisdiction.

39 Care should be exercised in interpreting rates based on small numbers of victims.

40 This release presents statistics about selected characteristics of victims of Family and Domestic Violence (FDV) -related offences which were reported to, or detected by police between 1 January and 31 December 2018. Additionally, selected statistics are available for the 2014 to 2017 reference periods. Users should be aware that movements in this data between reference periods may be reflective of changes in reporting behaviour or variances in the police detection that have occurred over the reference period.

41 These data are not designed to provide a total count of unique victims of FDV-related offences, nor the total number of individual FDV-related offences that came to the attention of police. For further information on the counting methodology refer to paragraphs 14–20.

42 Within the context of national Recorded Crime – Victims statistics, a Family and Domestic Violence (FDV) -related offence is defined as “An offence involving at least two persons who were in a specified family or domestic relationship at the time of the offence; or where the offence was determined by a police officer to be family and/or domestic violence related as part of their investigation”.

For the purposes of this release, a specified family or domestic relationship includes:

• Partner (spouse, husband, wife, boyfriend, and girlfriend)
• Ex-partner (Ex-spouse, ex-husband, ex-wife, ex- boyfriend, ex-girlfriend)
• Parent (including step-parents)
• Other family member (including, but not limited to, child, sibling, grandparent, aunt, uncle, cousin, niece, nephew)
• Other non-family member (carer, guardian, kinship relationships).

Selected offences are limited to the following ANZSOC subdivision offences:

• Murder
• Attempted Murder
• Manslaughter
• Assault
• Sexual assault
• Abduction and kidnapping.
   

43 The FDV-related data presented in this release have been derived using two variables, an FDV flag and Relationship of offender to victim (ROV). Where available, these variables are used to quality assure and further refine FDV-related data across the states and territories.

44 The FDV flag is one of a number of information sources used by police to determine whether an offence was FDV-related, and procedures for the use of the flag may vary across the jurisdictions as a result of legislative differences.

45 Police officers record the FDV flag when they have determined an offence or incident to be FDV-related as defined by the relevant state or territory legislation under which they operate. This differs across the states and territories:

• The FDV flag is recorded at the incident level in New South Wales, Victoria, Queensland, Tasmania, the Northern Territory and the Australian Capital Territory. This may result in a small number of victims in these jurisdictions being flagged as FDV-related which do not meet the definition of a specified FDV relationship. See paragraph 42.
• The FDV flag is recorded at the victim level in South Australia and Western Australia. This means that the flag is applied separately to each victim record and only selected offences which occurred within an FDV relationship, as specified in state legislation, are flagged.
   

46 For the purposes of improving data quality, known relationship information (ROV) was drawn upon to identify victims who had experienced victimisation within a specified family relationship (as outlined in paragraph 42) but were not flagged as FDV-related by police. These victims were re-classified as being FDV-related so that they were included in the FDV-related data presented in this release.

47 The ROV information provided for this collection is not detailed enough to separately identify those who experienced victimisation within a ‘carer’ and/or ‘kinship’ relationship as they are grouped with other relationships in the ‘other non-family member’ ROV category. This meant that although they aligned with the definition of a family and domestic relationship, they could not be checked to ensure the FDV-flag had been applied correctly. Consequently, FDV-related data on these victims were only included in the data where flagged as FDV-related by police.

48 This release may include data that were flagged by police as FDV-related but may not directly align with the specified relationship categories. For example, this may occur where no relationship information was provided to, or recorded by, police at the time of report.

49 Where the appropriate level of detail for ROV was available, offences flagged as FDV-related but which occurred against a stranger were removed from the data.

50 Data for the states and territories were largely based on ROV for all reference periods except for Western Australia. Although ROV has been provided from 2016 onwards for Western Australia, the data were not of sufficient quality due to there being a high level of unknowns. Therefore the FDV flag was used for deriving the majority of FDV-related data in this jurisdiction.

51 Users should note that the comparability of the FDV flag and ROV data, used to derive these counts, are influenced by variances in availability, legislation, business rules, and differences in the crime recording systems across the states and territories.

There are some known variances in the recording of relationship of offender to victim data across the states and territories which may impact on comparability of data in this release (see paragraphs 29–32).

52 Users should be aware that for this release, the utilisation of the FDV flag as a measure of victims of FDV-related offences varies across the states and territories as follows:

• Western Australian FDV data were based on the FDV flag and limited ROV information from 2016 onward. Prior to 2016, data are based solely on the FDV flag as recorded by police, and therefore may exclude victims of selected offences which occurred within a specified family or domestic relationship where these had not been flagged by police. As such, users are advised not to make direct comparisons across years.
• Tasmanian information about victims of FDV-related offences for 2014 to 2016 were based solely on the ROV data as recorded by police. From 2017, Tasmania Police have provided an FDV flag for the first time, however there was minimal impact on the data as the state specific legislation, under which the police flag is applied, only extends to include intimate partners in Tasmania.
   

53 The Census and Statistics Act 1905 provides the authority for the ABS to collect statistical information, and requires that statistical output shall not be published or disseminated in a manner that is likely to enable the identification of a particular person or organisation. This requirement means that the ABS must ensure that any statistical information about individuals cannot be derived from published data.

54 To minimise the risk of identifying individuals in aggregate statistics, a technique is used to randomly adjust cell values and summary variables. This technique is called perturbation and was applied to the Recorded Crime – Victims collection for the first time for the 2014 release. Perturbation involves small random adjustment of the statistics and is considered the most satisfactory technique for avoiding the release of identifiable statistics while maximising the range of information that can be released. These adjustments have a negligible impact on the underlying pattern of the statistics for the majority of this publication.

55 After perturbation, a given published cell value will be consistent across all tables. However, the sum of components of a total will not necessarily give the same result as the published total in a particular table. As such, proportions may add to more or less than 100%. Readers are advised to use the published totals rather than deriving totals based on the component cells.

56 Cells with relatively small values may be proportionally more affected by perturbation than large values. Users are advised against conducting analyses and drawing conclusions based on small values.

57 Perturbation has been applied to the majority of the data presented in this publication. However, the low levels of prevalence for Homicide and related offences, Kidnapping/abduction and Blackmail/extortion do not support the use of perturbation as an effective confidentiality technique for these data items. This issue, in combination with the high profile nature of Homicide offences, led to the decision to apply a different confidentiality process for these offences. Data have not been perturbed and some data have been suppressed to minimise the risk of identifying individuals in the aggregate statistics.

58 The 2010 publication marked a break in series for the Recorded Crime – Victims collection. This was due to changes in police recording practices, implementation of revisions to the ANZSOC and the implementation of the National Crime Recording Standard (NCRS). Data published prior to the 2010 collection are not strictly comparable

59 Statistics produced on the basis of date reported may be affected over time by lags in completing and/or processing some crime reports. Where offences reported in the reference year are not processed for inclusion in the national statistics until the following year, revised data are included in subsequent publications and noted accordingly. In this issue, 2017 data have been revised for New South Wales, Western Australia, Tasmania, the Northern Territory and the Australian Capital Territory to address this lag in reporting.

60 Given the complex nature of policing, many factors ultimately influence the level of recorded crime, which may not necessarily reflect changes in the actual number of criminal incidents:

• Social, cultural and economic factors may influence the level of criminal offending or the level of reporting to police.
• Recorded crime statistics are the by-product of an administrative system and are affected by changes within that system.
• The introduction of new technologies or changes in police business practices and resources are likely to influence levels of recorded crime.
• Changes to legislation may also have an impact on the level of recorded crime and on the types of offences recorded.
   

61 National statistics, therefore, require a level of uniformity when compiling data from different states and territories. Over time significant differences and changes in the business rules, procedures, systems, policies and recording practices of police agencies across Australia have resulted in some discrepancies in data between states and territories for some offence types. A number of standards, classifications and counting rules have been developed since the inception of this collection to improve national comparability.

62 In the early 2000s the National Crime Statistics Unit (NCSU) Board of Management commissioned the ABS to investigate discrepancies in the recording of crime statistics across jurisdictions.

63 Findings from the DiRCS project indicated that data for Assault were not comparable across all states and territories. Testing of this offence type highlighted that there were inconsistent recording practices across the states and territories. Some jurisdictions almost always recorded a reported criminal incident on their crime recording system, whereas other jurisdictions applied a threshold test prior to a record being made (e.g. whether the victim wished to proceed against the offender or the seriousness of the incident). These thresholds varied across jurisdictions and were not guided by national standards.

64 The project also concluded that once a crime had been recorded in a crime recording system, there was no evidence to suggest that processes within any state or territory had a significant impact on differences in recorded crime statistics.

65 In considering other aspects of recorded crime statistics, the DiRCS project and subsequent data quality investigations concluded that information for offence types other than Assault were satisfactory for the level of comparison presented in this publication. Where there are known specific issues for individual states and territories, these are described in paragraphs 75–91.

66 The NCRS was developed as part of the ABS's ongoing commitment to ensuring data comparability and quality. The NCRS complemented the already established classifications and counting rules for the Recorded Crime – Victims collection and improved the level of comparability of these statistics across jurisdictions.

67 The NCRS comprised a uniform set of business rules and requirements that were developed in collaboration with police agencies across Australia to guide the recording and counting of criminal incidents for statistical purposes and enable consistency in recording. A comprehensive set of scenarios has also been developed which underpin the rules and requirements of the NCRS. These scenarios provide police agencies with guidance about how to record an incident from the point at which it comes to police attention, to the point at which it is recorded into crime statistics.

68 The application of the rules and requirements of the NCRS enable the recording of crime for statistical purposes in a comparable manner, while still allowing for the recording and retaining of information on police systems for the primary reasons of operational investigation and law enforcement.

69 The application of rules and requirements of the NCRS was designed to enable the recording of crime in a comparable manner across all jurisdictions. However, there is some variability in the interpretation of the rules, in particular for Rule 2 of the NCRS for incidents of assault, which guides what is recorded on police systems when an incident is reported to police. According to this rule Police officers are required to take a report at 'face value' and record an incident on their crime recording system. An investigation will then follow to determine whether a crime has been committed at law.

70 As a consequence of the lack of data comparability for Assault and as part of an ongoing commitment to ensuring national data comparability, it has been determined that Assault data will only be published for those jurisdictions complying with the NCRS.

71 In relation to reporting of incidents of assault there are differences in interpretation and implementation of this rule across states and territories. New South Wales, South Australia, Tasmania, Northern Territory, Australian Capital Territory and Western Australia have interpreted and implemented the rule as required, and data for these jurisdictions are available. Tasmanian data are not published for the years prior to 2014, as the data for these years were not confirmed as being comparable with the other jurisdictions. Published Assault data are suitable for cross-jurisdictional comparisons.

72 The interpretation and implementation of the recording rule for assault incidents varies from the standard for Victoria and Queensland. The variation is particularly around whether an element of investigation occurs before deciding to record an incident on their crime recording system. The degree to which this approach is taken and the potential impact on data across the jurisdictions varies and is discussed below under ‘State / Territory events and specific issues’.

73 National data for Assault are not available in the Recorded Crime – Victims publication. The Crime Victimisation Survey and Personal Safety Survey are national ABS surveys that collect information about experiences of assault from households across all states and territories in a consistent manner, and are thus able to produce national estimates of the prevalence of assault in Australia. For more information, refer to Crime Victimisation, Australia (cat. no. 4539.0) and Personal Safety, Australia (4906.0).

74 The following information highlights events or processes unique to a state/territory that may have an impact on the data for this collection. This may include specific initiatives, recording practices, or changes to legislation or policy supplied by each state/territory. Where differences between jurisdictions relate to specific data items, they are discussed in paragraphs 23–32.

75 The proportion of victims with an unknown Indigenous status in New South Wales has continued to increase across all offence types over the past several reference periods due to recording at the operational level. Care should be taken when comparing Indigenous status data across reference periods.

76 Outcome of investigation data for the offence category of Other theft are unavailable for the years 2010–2013 due to coding issues for this data item. Consequently national data have been suppressed.

77 Prior to 2012, NSW age data were based on the age of victim at the date the incident occurred. From 2012 onwards, NSW age data are based on age of victim at the date an offence is reported to police in accordance with all other jurisdictions.

78 Counts of Kidnapping/abduction may be slightly inflated. 'Deprivation of liberty' (which is out of scope for this collection) is not separately identifiable on the NSW Police recording system; therefore counts of this offence type are also included in the Kidnapping/abduction offence category.

79 Additional codes were added to the Victoria Police Law Enforcement Assistance Program (LEAP) computer system in July 2015 to support finer disaggregation of relationship type. Prior to this system update, the relationship type of parent and child were grouped together in the same code when recorded in LEAP and were then disaggregated into the separate categories by Victoria Police when the data were provided to ABS. Caution should be used when comparing relationship data within these categories prior to 2015.

80 During the 2014 collection cycle, it was discovered that the national counting rules had not been applied correctly to Victorian data in previous reference years, resulting in an overstatement of the number of victims for most offence types. The statistical impact of this error was found to be negligible for most offence types, with the exception of Sexual assault and Motor vehicle theft. Consequently, data were revised for 2013 and users are advised that published data from 2010–2012 are not directly comparable with data published for 2013 onwards.

81 Property taken in association with UEWI may not always be identified due to limitations in recording options in the Victoria Police LEAP computer system. Therefore, caution should be used when assessing the sub components of UEWI. The total counts for UEWI are correct, however further disaggregation results in an undercount for 'UEWI – involving the taking of property' and an over count of 'UEWI – other'.

82 For incidents of Assault, Victoria differs from other jurisdictions in the interpretation and implementation of the NCRS (see paragraphs 66–73 for more information). Some element of an investigation will be undertaken before deciding whether to record an incident on their crime recording system. A record of the incident may be taken on other systems however the incident will not be recorded on the police recording system until it is determined that a crime has been committed. As a result of the comparability issues arising from this difference in interpretation and implementation of the NCRS, a decision has been made not to make available Assault data for Victoria.

83 Where an incident involving Assault is reported to police, it is not taken on face value and recorded on the Queensland Police Records and Information Management Exchange (QPRIME). Where a domestic violence incident occurs which involves an alleged Assault and the victim does not consent to proceeding with an assault charge, the assault matter is not recorded on QPRIME. In other jurisdictions, the incident would be included in the police recorded crime data even if the victim does not want to proceed with the assault investigation. As a result of the comparability issues arising from this difference in the interpretation and implementation of the NCRS a decision has been made not to make available Assault data for Queensland.

84 South Australia does not record an Assault for an unknown victim but may record another offence such as theft (when evidence suggests an assault or another offence has taken place) if they cannot locate a victim or their representative.

85 South Australia Police record a single victim in instances where multiple vehicles belonging to the same owner are stolen in a single incident. Victims of motor vehicle theft may therefore be understated. However, impact to victim counts is considered minimal.

86 The legal age of consent to sex for South Australia and Tasmania is 17 years of age. The legal age of consent in all other states and territories is 16 years.

87 There was an undercount of victims of Sexual assault in Western Australia prior to 2017. A review of coding identified offences which should be coded to 031 Aggravated sexual assault were coded to 032 Non-assaultive sexual offences (which is out of scope for the Recorded Crime - Victims collection). These offences have been correctly coded for the revised 2017 data as well as the new 2018 data. This should be considered when comparing the 2017 data onwards to the earlier reference periods.

88 Data users are advised to exercise caution when comparing Other theft data by location from 2017 onwards and earlier reference periods. Improved coding by Western Australia Police resulted in two changes in the data from the 2017 reference period onwards:

• An increase in the number of Other theft victims coded to ‘Outbuilding/residential land’. These victims had previously been coded to ‘Dwelling – private’ or ‘Dwelling – non-private’.
• An increase in the number of Other theft victims coded to ‘Service station’. These victims had previously been coded to ‘Retail n.e.c’. This location grouping detail was not published as its own category but was included in the ‘Retail’ category.

These coding changes affect both the Western Australia and National Other theft data by location. Users are advised to refer to the broader location groupings of ‘Residential’ and ‘Retail’ if comparing data prior to 2017.

89 The legal age of consent to sex for South Australia and Tasmania is 17 years of age. The legal age of consent in all other states and territories is 16 years.

90 The Northern Territory case management system does not include a 'no crime' outcome of investigation code. As a result, Northern Territory data may include victim counts for those situations where police have determined after investigation that 'no crime' has occurred. This differs to all other states and territories where 'no crime' data have been excluded from the victim counts.

91 Caution should be exercised when making any direct comparisons between different crime and justice data sources, as different collection methodologies can yield different results. Further information on crime data measurement issues is available in the information paper Measuring Victims of Crime: A Guide to Using Administrative and Survey Data (cat. no. 4522.0.55.001). This paper outlines differences between administrative data sourced from police agencies and survey data obtained directly from Australian households, and provides information to assist users with making informed decisions about which crime victimisation data source best meets their particular needs.

92 There are strong links between victims and offenders recorded by police in their administrative systems. Once a victim is recorded by police an investigation may ensue which could result, although not always, in an offender being proceeded against by police. However, there are a number of limitations in comparing the Recorded Crime – Victims collection and the Recorded Crime – Offenders collection:

• Data cannot be directly linked.
• Counting units vary as the concept of a principal offence is not applied in the Victims collection. Victims may be counted more than once if multiple offences reside in different ANZSOC divisions, or if there are multiple occurrences of victimisation in the reference period.
• The reference period used in the Victims collection is based on the calendar year, while the Offenders collection is based on the financial year.
• Police may detect a crime without it being reported by a victim. Additionally, ‘victimless’ crimes, such as Illicit drug offences or Regulatory offences are excluded from the Victims collection.
• Statistics about victims of Assault are not comparable across all states and territories. As a consequence, national data are not published.

Despite these differences, broad comparisons can be made between the two collections. For more information refer to Recorded Crime – Offenders, Australia (cat. no. 4519.0).

93 The ABS Crime Victimisation Survey collects information via personal telephone interview from persons aged 15 years and over about their experiences of crime victimisation in the previous 12 month period, for a range of selected personal and household crimes. The survey also collects information about selected sociodemographic characteristics of persons experiencing crime, whether they reported the most recent incident to police, and other selected characteristics of the most recent incident experienced. Results are published annually in Crime Victimisation, Australia (cat. no. 4530.0).

94 The ABS Personal Safety Survey collects information via face-to-face interview from men and women aged 18 years and over about the nature, extent, and characteristics of violence experienced since the age of 15. The survey also collects detailed information about experiences of partner violence and emotional abuse, stalking, sexual harassment, experiences of abuse before the age of 15, witnessing violence before the age of 15, and general feelings of safety. The survey is conducted every four years, with results from the most recent 2016 iteration of the survey published in Personal Safety, Australia, 2016 (cat. no. 4906.0).

95 The number of victims of murder and manslaughter published in Recorded Crime – Victims differ from the number of deaths recorded in Causes of Death, Australia (cat. no. 3303.0) as Assault (X85–Y09, Y87.1) i.e. murder, manslaughter and their sequelae. Reasons for the different counts include differences in scope and coverage between the two collections.

96 Recorded Crime – Victims statistics are compiled on a victim basis in that they count the number of victims for each individual offence category, rather than the number of breaches of the criminal law. The statistics presented in this publication may be different from those published by police forces in individual states and territories. Different definitions of offences (see Glossary) and counting methodology (see paragraphs 14–20) will result in variations.

97 Current publications and other products released by the ABS are available from the ABS website www.abs.gov.au. The ABS also provides a Release Calendar on the website detailing products to be released in the next six months. The National Centre for Crime and Justice Statistics can be contacted by email crime.justice@abs.gov.au.

98 Non-ABS sources which may be of interest include:

• Australian Capital Territory Policing, Annual Report: https://www.police.act.gov.au/about-us/publications
• Australian Criminal Intelligence Commission: https://www.acic.gov.au/publications
• Australian Federal Police, Annual Report: https://www.afp.gov.au/about-us/publications-and-reports
• Australian Institute of Criminology: http://crimestats.aic.gov.au/
• Crime Statistics Agency Victoria, Crime statistics: http://www.crimestatistics.vic.gov.au/
• Northern Territory Police, Fire and Emergency Services: http://www.pfes.nt.gov.au/Police/Community-safety/nt-crime-statistics.aspx
• New South Wales Bureau of Crime Statistics and Research: http://www.bocsar.nsw.gov.au/
• New South Wales Police Force, Annual Report: https://www.police.nsw.gov.au/about_us/publications/publications/annual_report
• Queensland Police Service, Annual Statistical Review: https://www.police.qld.gov.au/corporatedocs/reportsPublications/default.htm
• Queensland Government Statistician’s Office, Crime Statistics and Research: http://www.qgso.qld.gov.au/about-us/our-services/crime-statistics-research/index.php
• South Australia Police: https://www.police.sa.gov.au/about-us/crime-statistics-map
• Tasmania Police, Annual Report: http://www.police.tas.gov.au/about-us/corporate-documents/
• Western Australia Police: https://www.police.wa.gov.au/Crime/CrimeStatistics#/start

The Australian Bureau of Statistics (ABS) maintains national collections on crime victimisation from two different sources: administrative records, such as data from state and territory police agencies; and household surveys. While these collections contain data about closely related concepts, in practice, survey data and administrative data are quite different.

At times, the results from different administrative and survey data sources may create divergent pictures of crime in the community. There is no single data source that is able to provide all the information required to build a full picture of crime victimisation in Australia. Rather, there are multiple sources of information which provide data on different aspects of victimisation.

Generally, the administrative data sources provide a view of crime victimisation that has been brought to the attention of the Criminal Justice System and may subsequently become the focus of investigation and other police actions. Generally, survey data sources provide greater content coverage and detail and are better suited to analysis that requires extensive use of demographic variables or a national level focus.

The following table summarises the key differences between two separate but complementary sources of ABS crime victimisation data: Recorded Crime – Victims, Australia (cat. no. 4510.0) and Crime Victimisation, Australia (cat. no. 4530.0).