Go to top of page


Assaults recorded by police

The ABS defines assault as

the direct (and immediate/confrontational) infliction of force, injury or violence upon a person or persons, or the direct (and immediate/confrontational) threat of force, injury or violence where there is an apprehension that the threat could be enacted (ABS 2012b).

In 2011, the ABS did not present aggregated national results for assault but rather a breakdown by state/territory was provided. However, data were only available for New South Wales, South Australia, the Northern Territory, Western Australia and the Australian Capital Territory in 2011. For these jurisdictions only, the total number of recorded assaults was 117,873. In the absence of available data for the other jurisdictions, these official statistics were not used for the current costing calculations. Official recorded crime statistics for attempted murder were, however, available nationally and indicated that there were 185 attempted murders in Australia in 2011 for persons of all age categories. This number was revised in 2012 in respect of the 2011 year (ABS 2013c).

Crime Victimisation Survey, Australia 2011–12

The CVS reported that in 2011–12, there were 1,093,400 victims of physical assault or threatened assault, aged 15 years and over (ABS 2013a). The CVS provided a breakdown of whether the individual had been physically assaulted and/or threatened with physical assault. In total, the CVS reported that 539,800 individuals had been physically assaulted and 667,700 individuals had been threatened with physical assault (either face-to-face or not face-to-face; ABS 2013a). The CVS also provided data for victims aged 18 years and over who had experienced physical assault as to whether they experienced physical injury and whether or not they had sought medical treatment. In total, 44.9 percent of individuals assaulted were physically injured and 55.1 percent not physically injured. Of those physically injured, 39.3 percent sought medical treatment (17.7% of all assault victims). The remaining 60.6 percent of those physically injured did not seek medical treatment (27.2% of all victims of assault). These proportions have been applied to the overall number of physical assaults, although age differences could not be accounted for due to data aggregation. No information was available on whether individuals were hospitalised, so to account for this, the proportion of individuals hospitalised was taken from the 1997–98 CSS (ABS 1999) and applied to the proportion injured and seeking medical treatment as reported in the CVS. It was necessary to use these estimates, as individuals requiring medical treatment did not necessarily require the treatment in hospital. In addition, the costs for individuals with injuries severe enough to require hospitalisation are far higher and the impact on the individual is much greater than if hospitalisation is not required. Data on hospitalisation were not available from more recent ABS PSS.

Estimating the number of assaults

There were 1,093,400 victims of assault aged 15 years and over reported in the CVS for 2011–12 (ABS 2013a). Mayhew (2003b) noted that crime victimisation survey counts of assault may be inflated relative to police figures, even taking account of non-reporting. One reason is that although threats are, in principle, included in police figures, the police probably take a conservative approach to recording all threats brought to their attention. Another reason concerns multiple victimisation, which can be very pronounced in relation to assault. Thus, while victims may say they have been victimised several times (by the same person in all probability), it is likely that if they go to the police, the police will not record every separate incident, especially if they believe that the incidents are closely related to each other (Mayhew 2003b). On balance and in the absence of more precise data, the present study used the number of victims of assault recorded in the CVS for 2011–12 without applying any multiplier or other correction.

In order to account for the number of assaults of victims aged under 15 years that are not captured by the CVS, a multiplier of 7.2 was applied. This was calculated by averaging the percentage of victims aged under 15 years out of all victims of assault identified in police recorded crime data for New South Wales, South Australia, Western Australia, Northern Territory and the Australian Capital Territory. In New South Wales, for example, there were 5,631 victims of assault aged under 15 years recorded by police in 2011 out of 71,169 assault victims of all ages (7.9%). The average of these percentages for the jurisdictions for which data were available was 7.2 percent. Applying this resulted in an estimated total number of victims of assault of all ages nationally of 1,172,333.

The additional 185 attempted murders from ABS Recorded Crime data resulted in a total of 1,172,518 that was used for the costing exercise.

By way of comparison with other police recorded crime data set out in Table 1, it is possible to estimate the number of assaults that might have been recorded by police nationally by dividing the CVS total (and attempted homicides) by a multiplier of 6.9. This yields an estimate of 169,903. This multiplier is based on the average of Mayhew’s 5.3 (2003b) and Rollings’ 5.2 (2008), adjusted to take account of the latest UK multiplier for common assault of 7.9 (Home Office 2011). This estimation will not be further used for the purposes of the costing calculations, and is only included to enable a comparison between police recorded crimes over the last decade (see Table 2).

Table 5 shows the total estimated assaults by injury type. As indicated above, the proportion of the total number of assaults resulting in injury and whether medical treatment was sought came from the 2011–12 CVS (ABS 2013a). The rate of hospitalisation was based on the proportion of victims of assault hospitalised in the 1997–98 CSS (ABS 1999).

Table 5 Estimated number of assaults
Category Percentage of all victims of assaulta Number of estimated victims of assaultb
Hospitalisedc 1.83 21,639d
Injured, medical treatment without hospitalisation 15.82 185,463
Injured, no medical treatment 27.21 318,992
No injury 55.13 646,307

a: Including victims under 15 years of age

b: Based on the proportions of respondents who reported being injured and whether they sought medical treatment in the 2011–12 CVS

c: This breakdown is based on the proportion of those injured who sought medical treatment in the 1997–98 CSS

d: It is assumed that all attempted murders required hospitalisation and so this number includes the 185 attempted murders

The estimated figure for the number of assaults resulting in a hospital stay can be validated against the number of hospitalisations where the principal diagnosis was assault. While 2011–12 figures were unavailable, there were approximately 24,500 hospitalisations where the principal diagnosis was assault in 2009–10 (Tovell et al. 2012). In 2011–12, hospitalisations due to assault, based on CVS data, were estimated to be 21,639. While the two figures are not directly comparable, hospitalisation data indicate that the current figure may be slightly underestimated.

Medical costs

Using the approach Mayhew (2003b) followed using data from the Australian study conducted by the MURAC (Watson & Ozanne-Smith 1997), the estimated medical costs of assault requiring hospitalisation was $11,600 in 2011 (see Table 6). The estimated cost of an injury requiring medical treatment without hospitalisation was $690. The average cost for treating an injured person medically was $1,800. Overall, the average cost per incident of assault was $320 and the total medical costs due to assault were $379m.

Lost output

In keeping with Mayhew’s (2003b) methodology, MURAC’s costs for lost output are applied to those who had medical treatment. Of those who were injured, but not medically treated, their lost output was set at a quarter of the cost of those individuals treated outside a hospital (eg in a doctor’s office). This accounts for loss of productivity through disruption to work or through having to take time off work. The UK estimate adopted by Mayhew (2003b) for assaults that did not result in injury was also used here and adjusted to 2011 prices.

The lost output for an assault, averaged across all incidents, was $1,200. It is clear, however, that incidents of assault that resulted in hospitalisation or medical treatment were clearly more expensive (see Table 6). Overall, the cost of lost output using this methodology was $1.44b.

Table 6 Costs of assaults—medical, lost output and intangible losses
Category Per incident cost ($) Total cost ($m)
Medical Lost output Intangible Medical Lost output Intangible
Hospitaliseda 11,600 32,300 13,100 251 699 283
Injured, medical treatment without hospitalisation 690 2,700 2,800 128 500 519
Total—injured, medical treatment 1800 5,700 3,900 379 1,199 802
Injured, no medical treatment - 670 670 - 214 214
Not injured - 40 400 - 26 259
Average per incident cost 320 1,200 1,100 - - -
Total - - - 379 1,439 1,275

a: Includes attempted murders

Intangible costs

In keeping with the methodology employed in Australian costs of crime reports (Mayhew 2003b; Rollings 2008; Smyth 2011), the ratio of BITRE’s intangible losses for non-fatal road accidents to their lost output figures was applied (BITRE 2010; 2000). The previous ratio for hospitalised injuries was able to be revised from 0.66 to 0.40 due to newly released data (see BITRE 2010). The remaining ratios were, however, unable to be revised as the data were not sufficiently disaggregated among the other categories and as such, the original ratios used in Mayhew (2003b) have been maintained.

Per incident, intangible losses for assault are estimated to be $1,100. This figure was considerably higher for those individuals who were injured and received some form of medical treatment, at $3,900 an incident. Overall, intangible losses due to physical assault totalled $1.28b.

Total costs

The estimated total cost for assault is $2,600 per incident and $3.03b overall (see Table 7). The largest component of assault costs was lost output, followed by the intangible losses, which differed from previous reports where the reverse was the case. This is likely due to the revised ratio of intangible losses to lost output (see above).

Table 7 Overall unit and total costs of assaults, 2011
Category Per incident cost ($) Total cost ($m)
Hospitalised 57,000 1,233
Injured, medical treatment without hospitalisation 6,100 1,147
Total injured, medical treatment 11,400 2,380
Injured, no medical treatment 1,300 428
Not injured 440 285
Average per incident cost 2,619 -
Total - 3,021
Last updated
3 November 2017