Skip to main content

Contrary to popular opinion, we have witnessed a fall in many types of crime over the past two decades. Between 1995 and 2013/14, all crime recorded by the Crime Survey for England and Wales fell 62%, with a 51% fall in robbery and a 17% fall in theft from the person. Despite widespread attention, there is still little consensus as to why we have seen such declines in crime. To make inferences about why crime has fallen, first we need to develop an in-depth understanding of the nature of specific offences over time.

Sparked by an interest in the crime drop and mobile phone theft, my research explores the changing nature of theft from the person and robbery of personal property over a 17-year period. I have focused on two elements of my research in this article – the type of goods stolen and the characteristics of theft and robbery victims.

Offenders know what they want to steal. They have a ‘clear repository of crime targets’1 both in terms of the victim and items stolen. In the case of theft and robbery between 1993 and 2010, cash and purses/wallets were consistently the two ‘hottest’ products (stolen in 55% and 51% of incidents on average), followed by credit/debit cards (26%) and mobile phones (19%).

This is not particularly surprising given that the majority of thieves steal to fund an immediate cash need. Much of what is stolen through theft and robbery is desirable to the legitimate buying public and many individuals will own and carry these items on their person. There was a decline in the theft of more ‘traditional’ products, for example chequebooks and documents. This evolution of hot products is likely to be driven by changes in the market for stolen goods.

Mobile phones showed the biggest increase, peaking in January-June 2004 at 35% of incidents. Handsets first appeared in the ‘top five’ stolen items in July-December 1999, coinciding with what was termed the ‘tipping point’ when over four million phones were sold in the last quarter of 1999.

From 1999 to 2004 there was a marked increase in mobile phone theft. The proportion of incidents where a phone was stolen, on the whole, began to decline from 2004. However, a report produced by the ONS documents an increase in handset theft in 2011, thought to be driven by the popularity of smartphones. With UK consumers identified as ‘early adopters of new technologies’ the impact of particular products on theft and robbery trends should not be underestimated.

{mbox:significance/graphs/thomapson-robbery-graph.jpg|width=630|height=348|caption=Click to enlarge|title=Proportion (%) of Theft from the person and robbery incidents where selected goods stolen by six-month period (January 1993-June 2010). Dotted lines refer to missing data when the CSEW was not a continuous survey. Source: THOMPSON, R., 2014. Understanding theft from the person and robbery of personal property victimisation trends in England and Wales, 1994-2010/11. PhD. Nottingham Trent University.}

In order to delve more deeply into the phenomenon of falling crime, this research also explored the characteristics of victims. Previous research suggests we can identify, with some level of precision, the probability of victimisation using information regarding people, places and time2.

A number of potential risk factors were analysed under three broad categories: demographic, lifestyle and area of residence. The analysis suggests that, across sweeps, there are a number of indicators that consistently influence theft and robbery incidence. In line with previous research, the most common are age, sex, marital status and frequency of activity outside the home (in particular nightclub visits).

In addition, an individual’s general health, housing tenure status and car ownership/use significantly impact upon theft and robbery victimisation. Specifically, with each year of age, there is a decrease in the predicted mean number of victimisations.

With regard to sex, males experience a much reduced predicted mean number when compared to an otherwise similar female. Being single (as opposed to married) increases the predicted mean number of victimisations by anywhere between 38% and 95%.

In terms of lifestyle, visiting a nightclub, owning a car and the number of hours spent away from the home on an average weekday are important. The greater the number of visits to a nightclub per month (compared to someone who does not go at all), the higher the predicted mean number of victimisations.

Conversely, car ownership or use in the last year (as opposed to no car) decreases victimisations by half and spending less than an hour per weekday away from home (compared to someone who spends three to five hours away from home) can reduce victimisations by nearly two-thirds (64%).

In agreement with much literature in this field, living in London (compared to the South East) dramatically increases the predicted mean number of victimisations (between 58% and 191%) as does living in an inner city (between 46% and 106%) as opposed to a rural area.

As mentioned, the intrinsic characteristics of an individual, the lifestyle they lead and the area they live in are all important in predicting victimisation incidence. These characteristics have remained relatively consistent over time. Hence, within the context of the crime drop, this suggests that the same types of individuals are being targeted but less often.

It is hoped this research can help inform subsequent theories regarding the fall in crime, as well as broader policy and local practices to reduce the number of victims3. The findings from this research may help in the cost-effective allocation of (increasingly scarce) resources. With regard to victims, the findings could be used to personalise prevention, inform patrolling patterns and/or alter the design of the environment to reduce theft and robbery.

With regard to policy and practice, there are three key messages to take from this research. Firstly, we must routinely and systematically gather detailed information about stolen goods and how the market for them operates. This type of information allows us not only to see which items were stolen but also anticipate which may be stolen in future. The Home Office’s recent production of a ‘Mobile Phone Theft Index‘ should be commended on this account.

Secondly, we must continue to drive efforts in ‘designing out’ theft from new products. Phone manufacturers have taken some recent steps to improve security. Hopefully, both practices will help in moving toward security becoming the default and away from a position where ‘the criminal opportunities offered by potential victims are an undesired side-effect of their possession of certain goods4.’

Finally, the reasons why particular goods, individuals and environments may be more vulnerable to theft must be explored in more detail in order to enhance resilience and ultimately reduce the number of victims of crime.

This research was funded by a Nottingham Trent University Vice Chancellor’s Scholarship. Rebecca’s thesis is available to download here.

Footnotes

  • 1. Jacobs, B.A. (2010) ‘Serendipity in Robbery Target Selection’, British Journal of Criminology, 50(3), pp. 514-529. DOI: 10.1093/bjc/azq002.
  • 2. Negative binomial regression models (Cameron and Trivedi, 1986; Osborn and Tseloni, 1998 [see below]) were used to model the full distribution of crime counts for each sweep of the Crime Survey for England and Wales from 1994 to 2010/11. From this, the predicted mean number of theft or robbery victimisations for an individual with particular characteristics can be calculated. The results presented here show the percentage change in the predicted mean number of victimisations compared to the reference individual/category.
  • 3. Hough, M., Maxfield, M., Morris, B. and Simmons, J. (2007) ‘The British Crime Survey After 25 Years: Progress, Problems, and Prospects’ in Hough, M. and Maxfield, M. (eds) Surveying Crime in the 21st Century: Commemorating the 25th Anniversary of the British Crime Survey, Crime Prevention Studies Volume 22, Devon: Willan Publishing, pp. 7-31.
  • 4. Van Dijk, J.J.M. (1994) ‘Understanding Crime Rates: on the Interactions between the Rational Choices of Victims and Offenders’, British Journal of Criminology, 34, pp. 105-121.
  • Cameron, C.A. and Trivedi, P.K. (1986) ‘Econometric Models Based on Count Data: Comparisons and Applications of Some Estimators and Tests’, Journal of Applied Econometrics, 1, pp. 29-53. DOI: 10.1002/jae.3950010104.
  • Osborn, D.R. and Tseloni, A. (1998) ‘The Distribution of Household Property Crimes’, Journal of Quantitative Criminology, 14(3), pp. 307-330. DOI: 10.1023/A:1023086530548.

Leave a Reply