Australian Institute of Criminology

Skip to content

Thefts from vehicles

Estimating the number of thefts from vehicles

Vehicle thefts recorded by police

ABS Recorded Crime—Victims Australia 2011 does not report specific data on ‘theft from vehicles’ but rather this type of theft is included under ‘other thefts’ along with theft from a person (excluding the use of force), theft from a retail premises, theft not elsewhere covered and illegal use of property (except motor vehicles) (ABS 2012b). As a result, no specific police recorded crime data were available on theft from a motor vehicle from the ABS. Instead, data were available from a number of police jurisdictions that recorded incidents in their annual reports or police crime statistics. New South Wales, Victoria, Queensland, South Australia, Tasmania and the Northern Territory all provided data on the number of thefts from vehicles, which totalled 139,993 for 2011–12. Missing data from the Australian Capital Territory and Western Australia were estimated by calculating the number of other thefts that occurred in these two jurisdictions as a proportion of all other thefts and applying the same proportion to gross-up the total from the remaining jurisdictions. This resulted in a national estimated total of 168,666 for all states and territories in Australia.

Theft from vehicles in the Crime Victimisation Australia survey

As with many other types of crime, there is a certain level of under-reporting of thefts from vehicles. The ABS CVS 2011–12 (ABS 2013a) included questions about any incidents of theft from a motor vehicle that may have occurred to any member of their household in the previous 12 months. The CVS only included household members aged 15 years and older and ideally, it would have been appropriate to have inflated the number of incidents to account for victims under 15 years of age. However, owing to the fact that the police recorded crime statistics were not disaggregated by age categories, it was impossible to inflate the CVS data by the proportion of police recorded incidents of thefts from vehicles that involved persons under 15 years of age. The CVS defined theft from a motor vehicle as involving

the theft of property owned by any member of the household from a motor vehicle owned (for private use) by any member of that household (ABS 2013a: np).

The participants were asked to exclude property that belonged to someone from outside the household, any property that belonged to a business or employer, or property stolen from commercial vehicles. Also excluded were attempted break-ins to a vehicle if no property was stolen.

During the 12 months prior to interview, respondents indicated that they had experienced 379,200 incidents of theft of property from a motor vehicle (ABS 2013a). Around half (51%) of victims of theft of property from a motor vehicle had reported to police the most recent incident they had experienced. When the 2008 AIC report on costs of crime was undertaken, data were unavailable from the CSS (ABS 2006a) on theft from a motor vehicle. At that time, Rollings (2008) estimated the number of thefts from motor vehicles using UK figures, which indicated that slightly more than one-third of thefts from motor vehicles were reported to police, producing a multiplier of 2.8 (Dubourg, Hamed & Thorns 2005: 10). The UK Home Office (2011) recently updated that multiplier based on the British Crime Survey and recorded crime data to 3.5 for thefts from motor vehicles. For present purposes, however, it is safe to rely on the actual number of thefts from motor vehicles reported by respondents to the 2011–12 CVS (n=379,200), which produces a multiplier of 2.3 times the number of officially recorded thefts.

Estimating property loss

In order to estimate the value of losses arising from theft from motor vehicles, it is necessary to distinguish between thefts from private and commercial vehicles, as different mean losses apply to each of these categories. Rollings (2008) presented data from New South Wales that indicated that 15 percent of thefts from motor vehicles involved commercial vehicles and 85 percent related to private vehicles. In the absence of more recent data, this would mean that of the 379,200 thefts from motor vehicles in 2011–12, 56,880 would have involved commercial vehicles and 322,320 private vehicles.

Arguably, it would also have been beneficial to have estimated the value of losses arising from thefts from motor vehicles that were not stolen and theft from motor vehicles that were stolen. Estimates of the latter category could also be made for thefts from vehicles that had been recovered and thefts from vehicles not recovered as the unit cost is likely to be different for these various situations. Unfortunately, the available data did not permit such disaggregation of the results.

Updated estimates of property loss were available for only two state police jurisdictions, New South Wales for 2007 (Varshney & Fitzgerald 2008) and Victoria for 2011 (Victoria Police Crime Statistics 2011). These figures are likely to represent the higher end of the spectrum of crimes, as high-value crimes are more likely to be reported. In New South Wales, research was conducted by the New South Wales Bureau of Crime Statistics and Research into 150 randomly selected incidents of theft from a motor vehicle that occurred in the first six months of 2007. It was found that the average cost of a theft from a private vehicle was $603 in 2007 (Varshney & Fitzgerald 2008); which in 2011 prices would equate to $680 (RBA 2013). Varshney and Fitzgerald (2008) found that money was the most frequently stolen item from a private vehicle, followed by the vehicle number plates; however, the most costly item stolen from vehicles was laptop computers, with a mean loss of $2,333.

Victorian police statistics (Victorian Police 2011) reported an average value lost from theft from any type of motor vehicles of $805 for 2011, somewhat higher than the New South Wales estimate.

Earlier estimates from the United Kingdom showed a mean loss of $685 for individuals (Dubourg, Hamed & Thorns 2005) and estimates from the Crime Against Retail and Manufacturing Premises Survey showed a median loss of $571 for retailers per incident and $713 for each incident for manufacturers (Shury et al. 2005b). Mayhew (2003b) estimated costs of $600 for a theft from a commercial vehicle and $250 for theft from a non-commercial vehicle, which in 2011 prices would be $775 for a theft from a commercial vehicle and $323 for a theft from a private vehicle (RBA 2013).

Rollings (2008) estimated the cost of a theft from a commercial vehicle to be $555 and theft from a private vehicle to be $315, which inflated to 2011 prices would be $664 (commercial) and $377 (private). Rollings’ (2008) inflated cost of theft from a commercial vehicle seems reasonable and in line with Mayhew’s (2003b) estimate in 2001 and other estimates. In the case of theft from private vehicles, there are wide variations in previous estimates from $200 in New South Wales in 2005 to $685 in the United Kingdom in 2005. Arguably, the most recent and authoritative estimate is that provided by Varshney and Fitzgerald (2008) of $603 for New South Wales in 2007, which in 2011 prices would be $680.

Using Rollings’ (2008) mean cost of theft from a commercial vehicle, inflated to 2011 prices of $664, and Varshney and Fitzgerald’s (2008) cost of a theft from a private vehicle inflated to 2011 prices of $680, it is possible to obtain a total cost of thefts from all vehicles of $256.9m (see Table 18).

Damage caused in connection with thefts from vehicles

Varshney and Fitzgerald (2008) examined 150 incidents of theft from a vehicle and found that 35 percent of incidents involved a smashed window, a further 13 percent had locks tampered with or damaged and another seven percent involved some other structural damage to the vehicle. Varshney and Fitzgerald (2008) also found that in nine percent of thefts from vehicles, the vehicles were unlocked and no damage was sustained. Accordingly, it is appropriate to add 90 percent of the estimated damage costs when calculating the total cost of theft from motor vehicles.

The property damage amount per incident was derived from Dubourg, Hamed and Thorns (2005) who estimated that in 2003, £126 was recovered from each theft from a vehicle. This figure was inflated to 2011 UK prices and converted to Australian dollars using the OECD PPP for 2011, which resulted in an amount of $359. As no equivalent figures could be found for damage incurred from theft of a commercial vehicle, the $359 has been applied to both commercial and private vehicles. The total cost of damage caused in connection with thefts from motor vehicles is $136m (see Table 19).

Table 19 Costs of thefts from vehicles—property loss, lost output and intangible lossesa
Category Per incident cost ($) Total cost ($m)b
Property loss Damage Lost output Intangible Property loss Damage Lost output Intangible
Commercial vehicles 664 359 65 759 38 20 4 43
Private vehicles 680 359 57 759 219 116 18 245
Average per incident cost 678 359 58 759 - - -
Total -   - - 257 136 22 288

a: Medical costs are not estimated

b: Totals may not add to sub-components due to rounding

Lost output

The UK research into crime against retail and manufacturing premises conducted as part of the 2002 Commercial Victimisation Survey estimated that a median of two working hours were spent dealing with theft from vehicles in a retail setting and three hours spent dealing with such theft in a manufacturing setting (Shury et al. 2005b). Taking an approximate number of hours spent as being 2.5 and applying Australian average hourly rate in 2011 of $25.83 gives an estimated lost output of $64.60 per incident for theft from a commercial vehicle in Australia. The UK estimate for crimes against individuals and households estimated the lost output from each incident of theft from a vehicle as being the equivalent of $57; this figure being derived by inflating the 2003 cost in the UK report, inflating to 2011 UK prices and then converting to Australian dollars based on the OECDs pricing and PPP rate of $A2.20 to £1.00 (Dubourg, Hamed & Thorns 2005). Applying these estimates to the number of offences gives an estimate of $22m for lost output arising from theft from vehicles in Australia in 2011 (see Table18).

Intangible costs

The estimate for intangible costs from the UK 2002 Commercial Victimisation Survey report has been used in the absence of Australian data. As with all figures presented in this paper, the figure used was based on the UK price in 2003 inflated to 2011 UK prices and then converted to Australian dollars using the OECDs PPP as detailed above. This gave an estimated $759 per incident for intangible costs, which has been applied to thefts from commercial and private vehicles, giving a total of $288m for intangible costs.

Recoveries

Dubourg, Hamed and Thorns (2005) estimated that for each theft from a private vehicle, £11 worth of property was recovered. Inflating that amount to 2011 prices and converting to Australian dollars using the OECD PPP conversion results in a recovered property amount of $31 per incident. There were no equivalent Australian figures available and the amount of property recovered was only available for private vehicles. Nonetheless, it is reasonable to assume that recoveries from commercial vehicles would be similar to those from private vehicles and so the same recovery amount of $31 per incident has been applied to all thefts from vehicles. Applying this amount to the 379,200 thefts from vehicles in 2011 produces a total amount recovered of $11.8m.

Total costs

The total cost of thefts from vehicles is estimated to be $677m, or $1,785 per incident (see Table 20). The estimated cost of actual property loss and damage, less the value of goods recovered, represents slightly more than half of the total.

Table 20 Overall unit and total costs of theft from vehiclesa
Category Per incident cost ($) Total cost ($m)
Property loss 636 257
Damage 359 122
Lost output 58 22
Intangible 759 288
Sub-total 1,813 689
Less recovered property 31 12
Total 1,785 677

a: Medical costs are not estimated