Defining shop theft
The ABS category of ‘theft and related offences’ is defined as:
The unlawful taking or obtaining of money or goods, not involving the use of force, threat of force or violence, coercion or deception, with the intent to permanently or temporarily deprive the owner or possessor of the use of the money or goods, or the receiving or handling of money or goods obtained unlawfully (ABS 2011a: 52).
For present purposes, shop theft occurs where the victim of theft is a retail shop and includes the theft of goods, other than motor vehicles, by failing to pay or by avoiding payment for the goods obtained. Examples include shoplifting, theft from market stalls, theft from factory retail outlets and theft by employees of retail premises (ABS 2011a). Theft of motor vehicles, burglary and robbery at non-residential locations are excluded for present purposes and are dealt with in other sections of this report.
Estimating the number of shop thefts
ABS Recorded crime—Victims (2012b) shows that police recorded 487,573 victims of ‘other theft’ in 2011. Separate data are not provided by ABS for thefts from retail shops. However, some state police jurisdictions specifically identify shop theft (which is theft from retail premises by staff or customers) in their annual reports. For the 2011–12 financial year, New South Wales (NSW Police Force 2012), Queensland (QPS 2012) and South Australia (SA Police 2012) recorded 45,144 incidents of shop theft. Victoria Police (2011) recorded 19,356 incidents of ‘shop-steal’ for the period 1 January 2011 to 31 December 2011. Combined, these states recorded 64,500 incidents of shop theft. In order to estimate the number of incidents of recorded shop theft that occurred in Tasmania, Western Australia, the Australian Capital Territory and the Northern Territory, the same proportion of incidents of ‘other theft’ from these jurisdictions reported in ABS Recorded Crime—Victims 2011 (ABS 2012b) was used to estimate the number of incidents of shop theft. It was found that 20 percent of incidents of ‘other theft’ were recorded by police in Tasmania, Western Australia, the Australian Capital Territory and the Northern Territory. Inflating the 64,500 estimated incidents of shop theft from New South Wales, Victoria, Queensland and South Australia by 20 percent results in a national estimated number of 80,625 incidents of shop theft that would have been recorded by police for 2011. Of course, this is an approximate estimate only as the category of other theft includes theft from a retail store, theft from a person (excluding by force), theft of motor vehicle parts and contents, theft (excluding vehicle), theft not elsewhere classified and illegal use of property (except motor vehicles; ABS 2012b).
Prior crime victimisation surveys have found that many incidents of shop theft are both undetected and unreported. In the absence of Australian survey data, reliance was placed on surveys of shop theft in the United Kingdom. The British Retail Consortium (2012) Retail Crime Survey, for example, found that only 56 percent of customer thefts were detected and that only 59 percent of employee thefts were detected. Farrington’s (1999) review of shoplifting studies found that between one in 40 and one in 250 acts of shoplifting led to a conviction and it was estimated that between one in 100 and one in 1,000 shoplifting incidents were actually recorded by police (Farrington 1999). Using these data, Mayhew (2003b) adopted a relatively conservative multiplier of 100 for shop theft, which was again used by Rollings (2008).
However, in 2011, the Home Office substantially revised its multiplier for shoplifting offences down from 100 to 16.1 using an offender-based approach rather than Farrington’s (1999) methodology which was based on a limited number of self-reported shoplifting incidents. The revised methodology is in line with the method used to calculate the cost of shoplifting for use in the UK’s Drug Harm Index. The Home Office (2011: 4) describe their new approach as follows:
The volume of shoplifting incidents is the sum of the number of incidents by arrestees and the number of incidents by non-arrestees. The number of incidents committed by arrestees has been estimated using data from the 2005/06 Arrestee Survey (AS). The number of incidents committed by non-arrestees has been estimated using data from the 2003 Offending, Crime and Justice Survey (OCJS). The volume of shoplifting episodes has been adjusted to take account of co-offending using an estimated co-offending rate of 1.8 from Farrington (1999).
In the absence of local Australian research, the Home Office’s revised multiplier of 16.1 has been applied to the present estimated 80,625 incidents of shop theft in 2011, resulting in a total for Australia of 1,298,063 incidents (see Table 21).
|Estimated number of shop thefts recorded by the policea||80,625|
|Estimated number of shop thefts||1,298,063|
|Best estimate of value of theft per incident ($)||70|
|Best estimate of total property loss ($ million)||90.9|
a: Based on New South Wales, Victoria, Queensland and South Australia, aggregated to Australia-wide
Estimating property loss
Police estimates of the value of shop theft are likely to be skewed to the higher end of the scale, as higher value crimes are more likely to be reported. Taking this into consideration, where available, the median value of costs have been reported rather than the mean value, as mean values are likely to be highly skewed towards a small number of high-value crimes. One source of information on retail theft is the Global Retail Theft Barometer Survey 2010 (CRR 2010). The 2010 survey included 31 Australian retail business respondents from the Asia–Pacific region, who reported the cost of an average incident of retail theft by customers in the Asia–Pacific region of US$74.52. However, if the person responsible for the theft was an employee that amount rose to US$392.56 (CRR 2010). Another source of information on the cost of shop theft incidents was research by Shury et al. (2005b) who found the median value of direct financial losses of theft by customers from retail premises to be £35 in 2003 and the median value of theft by employees to be £125. In 2011, the Home Office (2011) updated its estimation of the unit cost to £124, although this includes not only the value of goods stolen but also costs in anticipation of shoplifting and the cost of responses.
Recent Australian data from Victoria Police (Victoria Police 2011) found the median value of 15,474 ‘shop steal’ offences reported to police in 2010–11 to be $70. Offences that were reported to police that had no value stated were excluded from the calculations (Victoria Police 2011). No other Australian police agencies reported information on the value of shop thefts. This most recent Victorian figure of $70 is slightly less than both Mayhew’s (2003b) estimate of $110 and Rollings’ (2008) estimate of $108 per incident. However, in the absence of other available data, the present study makes use of the Victorian figure of $70 per incident, which for the estimated total number of 1,298,063 incidents, totals $90,864,410 for all shop thefts across Australia in 2011. No allowance has been made for the recovery of goods, as prior studies have shown recoveries to be low (Farrington 1999).
Lost output arises principally from dealing with offenders and from managing stock losses (Mayhew 2003b). Shury et al. (2005b) estimated that the median length of time spent on a retail theft by a customer was one hour, which equated to approximately $25.83 per incident using the average Australian hourly wage in 2011 reported by the ABS Labour Force survey findings (ABS 2012c). For present purposes, lost output was calculated by multiplying the average hourly wage by the number of estimated incidents of shop theft. Previous reports calculated the amount differently. Mayhew (2003b), for example, used the mean cost of $10 per theft based on the indirect costs reported by businesses in the Small Business Crime Survey 1999 (Taylor & Mayhew 2002). Rollings (2008) estimated lost output by taking into account the UK estimates and CPI movement between 2001 and 2005, finding a reasonable estimate of the amount per incident due to shop theft to be $15. Arguably, Shury et al.’s (2005b) estimate based on $25.83 per incident is more accurate. Using this for the 1,298,063 estimated Australian incidents in 2011, gives a total estimated cost of $33,528,967 for lost output due to shop theft (see Table 22).
|Category||Per incident cost ($)||Total cost ($m)|
Note: Totals may not add to sub-components due to rounding
The total costs of shop theft are estimated at $124.4m, or $96 per incident (see Table 22). The largest component (73%) of shop theft was attributable to property loss.
The Australian Retailers Association (ARA 2013) estimated that the financial impact of retail theft in Australia was approximately $7.5b in 2013. This figure is based on three percent of the retail industry’s $243b annual income being lost to theft. By contrast, the present report estimates losses based on the estimated number of incidents of shop theft and their estimated mean value. The total cost of shop theft in Australia finding of $124.4m is arguably a more accurate estimate, although further research is needed in Australia to confirm both the multiplier used and the mean cost per incident.