Armed robbery: who commits it and why?


Armed robbery stories figure prominently in the media but official statistics suggest that armed robbery has actually declined markedly since the early 2000s. Reasons for this decline are debated, although it appears that target hardening in banks, a reduced domestic heroin market and possible changes in offender profile may all be influencing the observed trend. This paper focuses on the third of these explanations by providing an overview of the type of person who commits armed robbery. It describes offenders' backgrounds, motivations to offend and attitudes towards victims, and how armed robberies are undertaken. Ongoing research into these areas is important as it assists in the development of effective and up-to-date crime prevention measures.

In the 1990s, the number of known victims of armed robberies in Australia increased from around 5,000 per year to a peak of over 11,000 in 2001 (Figure 1). Between 2001 and 2005 the number dropped steadily to around 6,000 (ABS 2006). The number of banks robbed and the number of victims of armed robberies involving firearms have generally mirrored this pattern. It is unlikely that these statistics represent the full extent of armed robbery in this country, as these data reflect only offences reported to and recorded by police. Factors underlying these trends may include reductions in regional and metropolitan bank branches and bank security hardening, such as through the use of protective screens, more secure safes, limiting the amount of cash held by tellers and so on (Borzycki 2003). Recent evidence also points to a correlation between robbery and theft incidents and heroin use (Donnelly, Weatherburn & Chilvers, 2004). Falls in heroin availability since early 2001 which coincide with similar falls in the incidence of armed robbery indicate the significance of this pattern for the specific crime of armed robbery. As will be discussed later, a further trend observed in armed robbery is that traditional career robbers may be turning to other, more easily obtained sources of illicit income, such as through the armed robbery of hotels and clubs.

Figure 1 compares lifetime rates of substance use in the sample with rates reported among Australians aged 14 years or over, from the 2004 National Drug Strategy household survey (AIHW 2005). As expected, reported rates of substance use were markedly higher among participants than among the general population, with participants 41 times more likely to report heroin use, 32 times more likely to report injecting drug use, and at least 5 times more likely to report ecstasy, inhalant, amphetamine and cocaine use.


Figure 1: Trends in reported armed and unarmed robbery in Australia, 1995-2005


Understanding the decision making processes of both imprisoned and active offenders is one important way to reduce risk to victims, especially in terms of harm done and the likelihood of being a victim. Importantly, understanding what motivates offenders to commit crime can help in devising effective prevention measures. This paper outlines what is known of armed robbery, mostly gleaned from interviews with prisoners. While research based on interviews with active armed robbers can offer a different offender perspective, few such studies have been undertaken. Much of what is outlined in this paper, particularly the discussion around offender decision making processes, is drawn from material that is now 10 or more years old. This suggests that new research is needed in this area, especially in view of advancements in security technology and possible changes in the armed robber offender profile.

The careers of armed robbers can be seen as a progressive scale, with particular characteristics varying as more robbery offences are committed over time. Beginning as amateurs, offenders tend to be younger and less experienced (that is, they do not have extensive prior records) and tend to rob individuals. A large proportion of amateur armed robbers are not likely to go on to become regular robbery offenders. Those who do continue committing robbery offences will become more professional as their career continues. They will be older, become more likely to rob commercial establishments and have more extensive prior criminal records. Committing more offences as their career continues, offenders become more likely to commit more robberies, engage in better planning and be motivated by life's needs, rather than drugs (Gill 2001; Matthews 2002; Mouzos & Borzycki 2003). The common characteristic among all types of armed robbers (and violent property offenders in general) however, is that the offender is a stranger to the victim in the great majority of cases, in contrast to most violent crimes against persons (Indermaur 1995).

Sociodemographics of armed robbers

Armed robbers in Australia and overseas are overwhelmingly male, with few females engaging in the crime. Australian data indicate that around 10 percent of armed robbers are females (Borzycki 2006). A study of convicted armed robbers in the United Kingdom (Morrison & O'Donnell 1994) found that as few as one percent of armed robbers were female.

Most convicted armed robbers are less than 30 years of age (Borzycki 2006). There is a slight difference in age between those who rob banks and other armed robbers, the former being somewhat older (Nugent et al. 1989). Few armed robbers in Australia or overseas have completed high school education and fewer still are engaged in skilled employment. In one Australian study (Kapardis 1988), 75 percent of convicted armed robbers possessed no employment skills and one-third had worked at some stage as a labourer.

Convicted armed robbers in Australia and the UK are largely of European descent, most studies indicating that around 80 percent of robbers are from such a background (Kapardis 1988, Morrison & O'Donnell 1994, Nugent et al. 1989). In the United States of America the majority of offenders are described as Afro-American (Erickson 1996, Wright & Decker 1997): this difference has been linked to a large crack cocaine market in that country (Wright & Decker 1997).

Recidivism and progression into armed robbery

Continuing involvement in criminal activity, both robbery and other criminal offences, characterises an armed robber's way of life. According to a recent AIC survey of adult male prisoners in Australia, the Drug Use Careers of Offenders (DUCO) study (Makkai & Payne 2003), offenders typically commence regular robbery in their late teens to early twenties (Figure 2). They then graduate to armed robbery having tried many other forms of primarily nonviolent crime, such as break and enter, motor vehicle theft, receiving stolen goods, fraud and so on. It is of little surprise then that most convicted armed robbers have extensive prior criminal records. Overall in Australia, regular armed robbery offenders serving prison terms report that over their lifetime they have an average of seven armed robbery charges, five convictions and two prison sentences (AICDUCO male survey 2001 [computer file]). However, and as noted in the introduction, their actual level of armed robbery offending is often far greater than officially reported, with over two-thirds reporting that the police rarely or never find out about the times that they have committed armed robbery. This pattern matches victims of crime typically reporting a greater number of offences as having occurred than appear in official police statistics (AIC 2006).


Figure 2: Reported age at which adult male prisoners became regular armed robbery offenders (percent)


The motivation to offend

For the majority of armed robbers, the crime is primarily about funding a particular lifestyle, particularly one fuelled by illicit drugs. For a small group of entrenched armed robbery offenders, the motivation to commit armed robbery appears to be more about earning a regular (illicit) income; a means to pay bills and support a family (that is, more like a regular job). Table 1 summarises a selection of published data on this issue. In particular, it shows that between about one-fifth and one-third of offenders in the selected studies cite 'money for drugs' as a prime motivating factor for committing armed robbery.

Table 1: Motivations for committing robbery (percentage of all respondents)
Motivation Feeney (1986) Nugent et al. (1989) Gill (2000)
USA Australia UK
Source: n/a = no relevant data provided/available
Money for drugs 22 30 29
Money for food/shelter/family 11 18 6
Money to pay debts 7 6 n/a
Unemployed n/a 4 41
It's what offender did for a living n/a 5 n/a
Revenge/lost temper/angry 5 2 6
Drunk/high on drugs at time 7 n/a n/a
Urged to by friends 4 n/a 24
(n) (82) (110) (341)

Drug and alcohol use

Matthews (2002) found that the use of drugs by armed robbers during a robbery takes two major forms. The first is that the person takes drugs or consumes alcohol and then decides to carry out a robbery. The second is that they decide to do a robbery and then take drugs or consume a quantity of alcohol to help them get through it. Offenders who use drugs typically do so extensively. Almost two-thirds of offenders in one study (BOCSAR 1987) who were convicted of eight or more robbery offences were regular drug (mainly heroin) users. As noted in the introduction, this demonstrates that drug dependency is an important ingredient in robbery offending in Australia, and in particular in the pattern of offending of those with multiple robbery convictions. More recent work continues to support this (Donnelly, Weatherburn & Chilvers 2004).

While the majority of armed robbers cite getting money for drugs as a key motivating factor for their crimes, some armed robbers report that they do not engage in regular drug taking, particularly those engaged in high-risk armed robberies. For example, in one UK study (Walsh 1986) many more professional armed robbers than opportunists reported being sober and drug free during their robberies to obtain money or goods. Australian data appear to support the finding that there are two broad types of armed robbery offender, defined by their patterns of drug use (see Figure 3). For example, of 188 regular armed robbers in the DUCO study who responded to a question about drug use in the six months before they were in prison, 41 percent indicated that they did not use heroin at all, while 37 percent reported that they used heroin at least once a day (AICDUCO male survey, 2001 [computer file]). Thus, one type could be characterised as having engaged in regular, high use drug taking, and the other in little or no drug taking. This general pattern adds further interest to the recent downward trend in Australian armed robberies. For example, while it is clear that the number of armed robberies in each category has fallen since 2001 (Figure 1), coinciding with a period of heroin shortage, these falls are less dramatic for firearm robberies than for either non-firearm or unarmed robberies. This adds weight to the suggestion that professional robbers (who are more likely to use firearms - see below) may not be as drug dependent as other armed robbers.


Figure 3: Regular armed robbers' self-reported drug use in the six months before imprisonment (percent) [see attached PDF for chart]


Recent trends in Australia indicate that while methylamphetamine levels have stabilised since 2003, they still remain relatively high among all types of offenders (Mouzos, Smith & Hind 2006). At this stage it is too early to say what the effect of this has been on the specific behaviour of armed robbery offenders. However, given that the use of amphetamines and amphetamine type stimulants (ATS) has been found to be associated with more violent and impulsive behaviour (Makkai & Payne 2003), this may have important repercussions for armed robbery victims in the future.

Economic proceeds of armed robberies

As outlined above, getting money for drugs or living appears to be a key motivating factor among armed robbers. The most significant factor that differentiates professional robbers from other types of robbers appears to be the value of the takings - professional armed robberies often result in high rewards. In one Australian study (Mouzos & Borzycki 2003), the amount stolen by professional armed robbers ranged from a minimum of about $700 to as much as $20,000 in one incident. In contrast, the maximum amount netted in opportunistic and amateur armed robberies was about $1,500. Similar differences in takings have been observed overseas (Gill 2000; Matthews 1996; Morrison & O'Donnell 1994).

Data from the AIC's National Armed Robbery Monitoring Program (NARMP) (Borzycki 2006) indicate that higher reported average values are associated with firearm robberies (mean=$4,085) than with other types of weapons, such as knives (mean=$901) or syringes (mean=$893). Among a number of possible reasons that firearm robberies might result in higher average gains are that locations with high value property (like banks) may employ more security, and offenders may perceive that they need highly dangerous weapons to instil enough fear into victims to overcome the security measures. It is also possible that higher average gains are a product of better planning: more professional offenders who engage in substantial pre-offence planning and target properties with the potential for high takings may simply opt for such weapons (Borzycki 2006).

Planning an armed robbery

The amount of time and effort put into planning an armed robbery varies greatly. There are two extremes: at one extreme is the robber who spends little or no time planning and so does not research the target, does not wear a disguise, and makes no plans for escape; this type of robber does not usually choose their target in a calm, deliberate manner, but rather in a state of perceived desperation. The other extreme is the robber who spends a number of weeks planning for the robbery, thoroughly researches the target, considers in detail any security measures that may be present and, where possible, takes action to overcome them, and spends considerable time and effort organising disguises and escape plans (Matthews 1996; Wright & Decker 1997). These two extremes reflect the amateur/opportunistic and professional types of armed robbery offenders respectively.

The amount of time spent planning an armed robbery also varies by the type of target, and there is a close relationship between the amount of money available at these targets and the consideration given to robbing them (Erickson 1996). Planning the escape appears to be the most critical factor for offenders. Other things that armed robbers appear to consider in their escape plans include security measures (guard/police presence), the type of escape vehicle (for example, stolen or privately owned vehicle), and the escape route (BOCSAR 1987; Erickson 1996; Indermaur 1996).

Number of offenders

Most armed robberies in Australia, even of commercial establishments, are committed by sole offenders (Borzycki, Sakurai & Mouzos 2004). However, the number of offenders involved in an armed robbery generally varies with the type of weapon used. For example, in one Australian study (Mouzos & Borzycki 2003), 63 percent of firearm robberies involved two offenders. Firearms were also found to be used more often in professional armed robberies. Similar patterns have been observed in the UK (Gill 2001). The majority of knife and syringe robberies were also found to be committed by offenders acting alone (Mouzos & Borzycki 2003). In another Australian study (Kapardis 1988), the majority of offenders acting in pairs (84%) and gangs of three or more (80%) used firearms, while fewer of the loners (73%) did so.

Commercial targets

Commercial robbery targets comprise a wide range of settings from banks and building societies to hotels and clubs, service stations, supermarkets and restaurants. Perceived by offenders as one of the most profitable of targets, banks have in the past often accounted for a large proportion of armed robberies of commercial establishments in Australia (BOCSAR 1987, Nugent et al. 1989), although recent data suggest that service stations are now more likely to be victimised (Borzycki 2006). Like banks, cash-in-transit vehicles are highly lucrative targets but are much less frequently robbed. This is because they are high risk: guards have to be confronted directly, which poses great personal risk to offenders, and there are higher levels of security than found in most other non-bank targets (Gill 2001).

Weapon use in commercial targets

Although knives are the most frequently used weapon in armed robberies, where commercial establishments, particularly financial institutions and licensed premises are targeted, firearms are equally or more likely to be used (Borzycki 2006). This is probably because of the greater likelihood of victim compliance. It also appears that the type of firearm used, at least in banks, has changed over time (Borzycki 2003; Matthews 2002). Pistols were used more often than longarms in Australian bank robberies in the early to mid 1980s, but the reverse held true for the late 1980s. However, between 1998 and 2002 pistols were once more used in the majority of robberies involving firearms (Borzycki 2003), similar to trends observed for other violent crimes, like homicide (Mouzos & Rushforth 2003).

The most dramatic change in weapon use over time is among offender gangs (three or more offenders). In the 1980s almost all gangs used weapons during bank robberies, with pistols and longarms equally likely to be used. However, between 1998 and 2000, most bank robberies committed by gangs involved no weapon, with only around 10 percent and six percent involving pistols and longarms, respectively (Borzycki 2003). Borzycki suggests that this change is linked to a trend towards less planning (for example, no use of disguises) and the use of intimidation through large numbers of offenders rather than through the use of firearms.

Attitudes towards victims, violence and weapon use

Research evidence about armed robbers' attitudes to victims (both individuals and organisations) and the use of violence and weapons in the commission of their crimes is divided between that describing offenders as largely unconcerned about the impact of their crime on victims (Wright & Decker 1997) and that describing offenders as taking little pleasure in intimidating and injuring their victims (Gill 2000). Nevertheless, levels of violence used in armed robberies are not simply a function of offender inclination, but are also dependent on the reactions of victims and bystanders. Offenders typically view the use of violence as being about gaining control over an unpredictable situation and reducing the resistance of victims. The resistance of victims to the demands of offenders, or the attempt of victims or other bystanders to otherwise obstruct them, greatly increases the risk of violence (Indermaur 1995; Mouzos & Borzycki 2003). People who stand in the way of an armed robber and his/her main objectives (money and escape) run the highest risk of physical injury. While the threatened use of violence pervades all armed robberies, most research demonstrates that the actual use of violence by offenders and serious (physical) injuries sustained by victims are relatively uncommon. For example, Borzycki (2003) indicates that only around six percent of bank hold-ups in Australia between 1998 and 2002 resulted in any physical injuries to victims.

Injuries to victims

Australian research indicates that the number of physical injuries to victims caused by firearms is relatively small compared with other weapon types (Borzycki 2006). While firearms can result in far more serious injuries than other weapons, they are infrequently discharged. However, where victims resist armed robbers, physical injury is much more likely to occur. In one study (BOCSAR 1987), the victim had offered physical resistance in 67 percent of the armed robberies resulting in serious injury. This compared with 33 percent where there was no victim resistance.

The presence of multiple offenders may pose an increased risk of victim injury, particularly where alcohol and loaded firearms are involved. In one Australian study (Kapardis 1988), a victim was found to be four times more likely to be injured if a gang of three or more offenders was involved than with a lone offender. Similarly, in a UK study (Morrison & O'Donnell 1994), where targets had no physical barriers (such as screens) and there were multiple offenders there was an increased risk of victims experiencing physical injuries.

To summarise, Australian and overseas research indicates that a useful distinction can be made between two categories of armed robbers. The first comprises professional robbers who engage in considerable planning, target high risk commercial establishments, use firearms more frequently and work in groups. The second category consists of amateur/opportunistic robbers who do not spend much time planning, direct their efforts at a range of low risk targets, do not typically use firearms and work largely independently. Trends in armed robbery also suggest that the type of offender targeting high risk targets (like banks) and the manner in which these offenders conduct their crime may be changing.

Target hardening, through the use of security screens, safes, security guards and so on, has reduced the risk of robbery for certain sorts of targets, like banks but, at least according to the Australian evidence, whether this has been at the expense of displacing the crime into other areas and/or onto other targets is not entirely certain (see Cusson 1993, however). This lack of clarity derives in most part from the fact that crimes prevented, as against those committed, leave no evidence that they have been prevented (Clarke 1990), but also because little research in this area appears to have been conducted recently, particularly in Australia. The small number of (now ageing) studies that deal with the issue of the displacement of armed robbery are divided in their findings, some suggesting that situational crime prevention measures have had minimal displacement effects (Clarke 1990), while others indicate that such measures may have had a greater impact on offenders (BOCSAR 1987).

Despite this lack of evidence, it has been suggested recently (Mouzos & Borzycki 2003) that target hardening in banks may be causing traditional career robbers to turn to other, more easily obtained, sources of illicit income (for example, hotels and clubs) leaving only the more opportunistic and spontaneous robbers to victimise banks. Further research on this trend will be possible in the future through the NARMP as more data become available. More detailed offender based research will also be needed to understand why this may be occurring. However, if this hypothesis turns out to be true, it raises some significant issues for situational crime prevention. For instance, licensed premises may need to learn from the experiences of the banking industry and adopt some of the target hardening measures that have become commonplace in most banks. Furthermore, banks may increasingly be confronted by a new type of offender who could be described as being more desperate, drug dependent (probably intoxicated at the time of the offence) and volatile. At this stage it is too early to predict what the future consequences of greater amphetamine/ATS use will be on armed robbery offenders and their victims. However, where offenders are unpredictable and irrational and where their decision making is erratic (whether because of drug use or otherwise), the development of effective situational crime prevention measures is challenging.


All URLs were correct at July 2006

  • Australian Bureau of Statistics (ABS) 2000-2006. Recorded crime, victims, Australia. ABS cat. no. 4510.0. Canberra: ABS
  • Australian Institute of Criminology (AIC) 2006. Australian crime: facts and figures 2005. Canberra: AIC
  • Borzycki M 2006. Armed Robbery in Australia: 2004 National Armed Robbery Monitoring Program annual report. Research and public policy series no. 69. Canberra: AIC
  • Borzycki M 2003. Bank robbery in Australia. Trends & issues in crime and criminal justice no. 253
  • Borzycki M, Sakurai Y & Mouzos J 2004. Armed robbery in Australia: 2003 National Armed Robbery Program annual report. Research and public policy series no. 62. Canberra: AIC
  • Clarke RV 1990. Armed robbery interim report 2: target hardening of banks and displacement of robberies. Adelaide: National Police Research Unit
  • Cusson M 1993. A strategic analysis of crime: criminal tactics as responses to precriminal situations. In Clarke RV & Felson M (eds), Routine activity and rational choice. New Brunswick: Transaction Publishers: 295-304
  • Donnelly N, Weatherburn D & Chilvers M 2004. The impact of the Australian heroin shortage on robbery in NSW. Crime and justice statistics. Sydney: NSW Bureau of Crime Statistics and Research
  • Erickson R 1996. Armed robbers and their crimes. Seattle: Athena Research Corporation
  • Feeney F 1986. Robbers as decision-makers. In Clarke R et al. (eds), The reasoning criminal: rational choice perspectives on offending. New York: Springer
  • Gill M 2001. The craft of robbers of cash-in-transit vans: crime facilitators and the entrepreneurial approach. International journal of the sociology of law 29(3): 277-291
  • Gill M 2000. Commercial robbery. Great Britain: Blackstone
  • Indermaur D 1996. Reducing the opportunities for violence in robbery and property crime: the perspectives of offenders and victims. In Homel R (ed), The politics and practice of situational crime prevention. Crime prevention studies no. 5. Monsey NY: Criminal Justice Press: 133-157
  • Indermaur D 1995. Violent property crime. Annandale NSW: Federation Press
  • Kapardis A 1988. One hundred convicted armed robbers in Melbourne: myths and reality. In Challinger D (ed), Armed robbery. AIC seminar proceedings no. 26. Canberra: AIC: 37-49
  • Makkai T & Payne J 2003. Drugs and crime: a study of incarcerated male offenders. Research and public policy series no. 52. Canberra: AIC
  • Matthews R 2002. Armed robbery. Devon: Willan
  • Matthews R 1996. Armed robbery: two police responses. Crime detection and prevention series paper no. 78. London: Home Office Police Research Group
  • Morrison S & O'Donnell I 1994. Armed robbery: a study in London. Occasional paper no. 15. Oxford: Centre for Criminological Research
  • Mouzos J & Borzycki M 2003. An exploratory analysis of armed robbery in Australia. Technical and background paper series no. 7. Canberra: AIC
  • Mouzos J & Rushforth 2003. Firearm related deaths in Australia 1991-2001.Trends & issues in crime and criminal justice no. 269
  • Mouzos J, Smith L & Hind N 2006. Drug Use Monitoring in Australia: 2005 annual report on drug use among police detainees. Research and public policy series no. 70. Canberra: AIC
  • New South Wales. Bureau of Crime Statistics and Research (BOCSAR) 1987. Robbery: an analysis of robbery in New South Wales. Sydney: Bureau of Crime Statistics and Research
  • Nugent S et al. 1989. Armed robbery from an offender's perspective: implications for prevention. Canberra: AIC
  • Walsh D 1986. Heavy business: commercial burglary and robbery. London: Routledge & Kegan Paul
  • Wright R & Decker S 1997. Armed robbers in action: stickups and street culture. Boston: Northeastern University Press

About the authors

Dr Katie Willis is a research analyst in the crime reduction and review program at the Australian Institute of Criminology