Armed robbery in Australia : 2003 National Armed Robbery Monitoring Program annual report

Maria Borzycki, Yuka Sakurai and Jenny Mouzos
ISBN 0 642 53872 7 ; ISSN 1326-6004
Canberra: Australian Institute of Criminology: 2004
(Research and public policy series, no. 62)

Analyses of the NARMP

A total of 8,865 valid victim cases were examined for this report. Unfortunately all variables were not supplied for all cases, so there are differences in the numbers of cases contributing to the various analyses. Similarly, the level of detail provided by each jurisdiction varied, limiting the degree of specificity possible in some analyses. Where highly detailed information was extracted, case numbers are often small, thus care should be exercised in attempting to generalise those findings to circumstances beyond the actual cases examined.⁵

Armed robberies

Location of armed robbery

Over half (59%) of the reported armed robberies (that is, armed robbery victimisations) in Australia in 2003 occurred in NSW, whereas the smaller jurisdictions (NT, ACT and Tasmania) collectively accounted for a little less than two per cent (see Figure 1). It is impossible to assess if the spread of victimisations is exactly mirrored in the spread of actual armed robbery incidents because, as already noted, the database is in a victim-based format.⁶

Armed robberies take place in a variety of locations (see Table 1), although the largest proportion of victimisations occurred in retail settings (46%, or 4,106 cases). Two-thirds of these victimisations occurred in locations simply classified as 'retail',⁷ but of the remaining third where greater detail was available, the most often-victimised retail sites were service stations (9% of all locations; 20% of all retail locations). The next most common armed robbery location can collectively be called open public settings (3,750 cases, or 42%). This incorporates footpaths, transport-related and other open spaces accessible to the community. Of these cases, the majority of victimisations occurred on the street and footpath (n=2,157).

Table 1 : Location of armed robbery

Location	Number	%
Residential setting
Residential setting	527	6
Total	527	6
Open public settings
Recreational settings (parks, cinemas etc; excluding licensed premises)	771	9
Transport-related settings (car parks, public transport, train, bus stations etc)	636	7
Open spaces excluding street and footpath (vacant lots, bushland etc)	186	2
Street and footpath	2,157	24
Total	3,750	42
Other community settings
Education, health, religious, justice and other community settings	95	1
Administrative and professional settings	182	2
Wholesalers, warehouses, manufacturing and agricultural settings	32	<1
Total	309	4
Retail settings
Retail settings with no further detail	2,728	31
Banking and financial settings	223	3
Pharmacies and chemists	277	3
Service stations	819	9
Licensed premises (bottle shop, pub, hotel and night club)	59	1
Total	4,106	46
Other
Unspecified/other locations with no further detail, or not classified elsewhere	173	2
Total	173	2
Grand total	8,865	100
Note: Percentages may not total 100 because of rounding Source: Australian Institute of Criminology NARMP 2003 [computer file]

Armed robbery can be thought of as an impersonal crime, insofar as one would assume that victims apparently chosen at random would have no prior relationship with the robbery offender. Interestingly, around one in twenty armed robberies (6%) took place in residential settings. It seems unlikely that offenders would randomly target private, residential premises without some prior knowledge of the location or its residents. The victim and offender in these settings may have some knowledge of each other prior to the event, which would make armed robbery in residential settings at odds with robbery in other locations. This victimisation in homes is explored in greater detail below, in the discussion of relationships between offender and victim.

Weapons use in armed robbery

A total of 9,332 weapon descriptions were associated with the armed robbery victims recorded in the NARMP. This number exceeds the number of victims because some jurisdictions were able to supply information regarding multiple-weapon armed robberies.

Consistent with previous Australian research (e.g. Mouzos & Carcach 2001), knives were the most used weapons in armed robberies (4,513 or 48% of all weapons reported; see Table 2). Firearms made up a little over one-fifth of all weapons recorded in the NARMP (22%), whereas 'other weapons' constituted nearly one-quarter (24%). Only four per cent of reported weapons were syringes.

Table 2 : Weapons used in armed robbery

	Number	%
Weapon used (with no further detail)	2	<1
No weapon use reported	19	<1
Unknown	107	1
Total	128	2
Firearms
Firearm (with no further detail)	350	4
Shotgun	192	2
Rifle or airgun	73	1
Handgun	963	10
Sawn-off longarm	57	1
Replica firearm	31	<1
Other firearm (not classified elsewhere)	386	4
Total	2,052	22
Knives
Knife with no further detail	4,149	44
Dagger	3	<1
Kitchen knife	7	<1
Scissors	75	1
Pocket knife	25	<1
Screwdriver	215	2
Other knife (not classified elsewhere)	39	<1
Total	4,513	48
Syringes
Syringe	387	4
Total	387	4
Other weapons
Other weapon type (with no further detail)	413	4
Club, baton or stick	208	2
Rock, stone or brick	21	<1
Tool	217	2
Blunt instrument (not classified elsewhere)	81	1
Bottle	87	1
Chemical spray	12	<1
Explosive or bomb	3	<1
Machete or axe	190	2
Sledgehammer	62	1
Crowbar or metal pipe	124	1
Other weapon (not classified elsewhere)	834	9
Total	2,252	24
Grand total	9,332	100
Note: Multiple weapon types were listed in some cases, therefore number refers to the number of weapon types, not the number of victim cases. Percentages may not total 100 because of rounding. Source: Australian Institute of Criminology NARMP 2003 [computer file]

Table 3 : Weapon combinations used in armed robbery

Weapon combination	Number	%
No specific weapon types listed	262	3
Single firearm	1,812	20
Single knife	4,122	46
Single syringe	365	4
Single 'other weapon'	1,768	20
Multiple firearms	36	<1
Multiple knives	85	1
Multiple syringes	2	<1
Multiple 'other weapons'	115	1
Firearm and knife	66	1
Firearm and 'other weapon'	86	1
Knife and syringe	9	<1
Knife and other weapon	120	1
Syringe and 'other weapon'	9	<1
Firearm, knife and 'other weapon'	8	<1
Total	8,865	100
Note: Percentages may not total 100 because of rounding Source: Australian Institute of Criminology NARMP 2003 [computer file]

On average, 1.4 weapon types were employed per case (with a median score of one), suggesting that the majority of robberies involved only one weapon: analysis of the weapon types used in combination showed that single weapon types were employed in the vast majority (91%) of cases. Reports of robberies with multiple single weapon types (e.g. multiple firearms), or of different weapon types used in combination, were only a minority of cases (both 3% of all victims; see Table 3). This could be interpreted as meaning that offender(s) rarely use multiple weapons. It is more probable, however, that these data reflect the difficulty jurisdictions encountered when attempting to accurately extract information on multiple weapons, on multiple offenders, and/or to provide offender information when no offender(s) had been apprehended. These methodological issues are highly relevant, because until the reliability of data can be improved, it will be difficult to ascertain the number and nature of multiple weapon armed robberies.

Some jurisdictions were able to supply detailed weapon information, and analysis of these data highlight the seemingly opportunistic nature of the weapon choice of some offenders. Of the 1,005 'other weapons' described in detail, 59 per cent were tools (including machetes, axes and crowbars), and eleven per cent were objects that could be found at an outdoor site (bottles, rocks, stones or bricks). In contrast, the most commonly employed firearm was that which is most easily concealed but also (presumably) the most difficult to obtain because of the severe restrictions governing its ownership and the severe penalties imposed for its illegal use - the handgun (73% of the 1,316 firearms described in detail).

Detailed weapon information suggests that armed robberies in 2003 can be broadly categorised as either opportunistic or planned, at least in terms of the preparation potentially needed to obtain weapons.

Weapons use in different locations

Weapons were not used in uniform proportions over all robbery sites (Table 4). In keeping with overall findings, knives were most commonly used in most locations,⁸ but firearms were proportionally the most common weapons in recreational settings (42%), administrative and professional settings (44%), and banking and financial settings (49%). Although knives were the most common weapons in licensed premises (44%), nearly one in three robberies (32%) involved a firearm. Less than a quarter of service station and pharmacy robberies involved threat with firearms (22% and 17% respectively).

Table 4 : Weapons used in armed robbery, as percentage of each location

Location	Firearm	Knife	Syringe	Other weapon	Number
Residential settings
Residential setting	20	46	1	33	539
Open public settings
Recreational settings (parks, cinemas, etc; excluding licensed premises)	42	33	1	23	871
Transport related settings (car parks, public transport, train and bus stations, etc)	13	53	6	27	637
Open spaces excluding street and footpath(vacant lots, bushland, etc)	8	64	4	24	193
Street and footpath	11	56	4	29	2,175
Other community settings
Education, health, religious, justice and other community settings	17	53	3	27	98
Administrative and professional settings	44	29	1	26	196
Wholesalers, warehouses, manufacturing and agricultural settings	6	56	3	34	32
Retail settings
Retail settings with no further detail	26	48	6	20	2,864
Banking and financial settings	49	25	2	24	239
Pharmacies and chemists	22	58	7	14	297
Service stations	17	55	4	24	828
Licensed premises (bottle shop, pub, hotel and night club)	32	44	0	25	57
Unspecified and other locations with no further detail, or not classified elsewhere	22	50	4	24	178
Total (per cent)	22	49	4	24	9,204
Note: Multiple weapon types were listed in some cases, therefore number refers to the number of weapon types, not the number of victim cases. Percentages may not total 100 because of rounding. Source: Australian Institute of Criminology NARMP 2003 [computer file]

Firearms were not often used in street robberies (11%): knives made up the highest proportion (56%), with 'other weapons' employed in nearly three in ten street robberies (29%). This pattern is similar to that seen in transport related settings, but is at odds with the weapon use pattern seen in recreational settings, the other 'public location' category. This anomaly may be because the recreational settings category encompasses public spaces that are probably subject to active security practices (e.g. cinemas, sporting venues and caravan parks). Recreational location weapon use patterns did resemble those seen in other presumably more secure locations, such as banking and financial, and administrative and professional settings. It is possible that security measures necessitate more offender planning, which may be related to less opportunistic weapon choice, or that hardened security results in an escalation of weapon choice, from knives or other opportunistic weapons, to firearms. More detailed data and/or sufficient cases would allow for the modelling of these potential relationships.

Interestingly, service stations (retail locations known to engage in security measures), showed patterns of weapon use that most resemble street robberies: knives constituted the majority of weapons used, then 'other weapons', then firearms. This may be a function of the hours of operation of these sites: longer hours of operation - hours outside the times when one would expect numerous customers who could be potential witnesses - make them more accessible to opportunistic offenders. Opportunistic robbers may not consider the obstacles posed by hardened security, or might consider the risk posed by security outweighs the perceived high gains (cash and tobacco) associated with this location.

Temporal aspects of armed robbery

There was little variation in the number of armed robberies perpetrated on different days of the week: the proportion of all robberies ranged between twelve per cent (Wednesdays) and sixteen per cent (Mondays). In general, the majority of robberies occurred in the evening and early morning (52% in the hours between 6pm and 3am). However, the spread over the time of day at which those robberies occurred was not the same for all days (Table 5). Over thirty-five per cent of weekend robberies (for these purposes, Friday, Saturday and Sunday) occurred between 9pm and 3am, whereas the equivalent figure was around thirty per cent for weekdays. Although this percentage difference appears minimal, it does represent a large number of cases, as day and time data were available for all but twelve cases.

Table 5 : Time when armed robberies occurred, as percentage of day of the week

Time	Sun	Mon	Tues	Wed	Thurs	Fri	Sat
Midnight to 2.59am	17	10	10	11	10	14	15
3am to 5.59am	9	6	4	6	6	6	9
6am to 8.59am	4	6	5	4	5	5	4
9am to 11.59am	6	10	8	10	9	9	7
Midday to 2.59pm	10	14	13	15	11	11	9
3pm to 5.59pm	12	19	19	17	18	15	16
6pm to 8.59pm	19	17	19	16	20	17	18
9pm to 11.59pm	23	18	21	21	21	24	21
Number	1,372	1,387	1,208	1,096	1,163	1,308	1,318
Note: Percentages may not total 100 because of rounding. Source: Australian Institute of Criminology NARMP 2003 [computer file]

Monthly variation was only moderate (ranging from 7% of all robberies in September, to 10% in April), as was seasonal variation (23% of all robberies in spring, ranging to 27% in autumn).⁹ Slight differences in the spread of time of day were observed, however, as a function of the season (Table 6): nearly four in ten (38%) summer time robberies occurred between 9pm and 3am, with less than three in ten (28%) taking place between 3pm and 9pm. This pattern was reversed in the winter, with forty-two percent reported as occurring in the late afternoon/early evening (3 to 9pm), and twenty-eight per cent taking place between 9pm and 3am. This may be a function of victim availability (more people about during warmer summer evenings) and/or greater offender activity (either because of more potential victims, or because of the more pleasant weather).

Table 6 : Time when armed robberies occurred, as percentage of seasons

Time	Summer	Autumn	Winter	Spring
Midnight to 2.59am	16	12	10	13
3am to 5.59am	8	6	5	8
6am to 8.59am	6	5	4	4
9am to 11.59am	8	7	9	9
Midday to 2.59pm	13	11	12	11
3pm to 5.59pm	15	16	20	16
6pm to 8.59pm	13	19	22	18
9pm to 11.59pm	22	24	18	21
Number	2,195	2,432	2,221	2,004
Note: Percentages may not total 100 because of rounding. Source: Australian Institute of Criminology NARMP 2003 [computer file]

The time at which robberies occurred also varied with location (Table 7). Premises that one would assume would operate under conventional opening hours (i.e. 9am to 5pm), such as education and health facilities, administrative and professional settings, financial institutions, unspecified retailers and pharmacies, had the greatest proportion of victimisations in the afternoon (3 to 6pm; 25%, 42%, 34%, 21% and 37% respectively). This may reflect a conscious choice of time and location on the part of offenders, who possibly expect larger available cash floats later in the business day. In contrast, public space robberies (recreational and transport related settings, and the street and footpath) most often occurred in the late evening (9pm to midnight; 29%, 23% and 24% respectively). Service stations were most often robbed between midnight and 3am (28%). As discussed above, offenders might target service stations at a time when they expect there to be minimal witnesses and resistance to their crime, using weapons that are easily accessed and which can be anonymously and easily discarded, ignoring the security measures service stations often have in place.

Table 7 : Time when armed robberies occurred, as percentage of locations

Location	24:00 to 2:59	3:00 to 5.59	6:00 to 8.59	9:00 to 11:59	12:00 to 14.59	15:00 to 17:59	18:00 to 20:59	21:00 to 23.59	N
Residential setting	14	8	5	6	13	12	17	25	522
Recreational settings	15	7	4	10	9	9	17	29	769
Transport related settings	11	4	3	10	14	18	16	23	636
Open spaces excluding street and footpath	19	10	2	10	14	16	17	12	186
Street and footpath	18	7	4	7	9	13	17	24	2156
Education, health, religious, justice and other community settings	14	5	4	8	13	25	17	14	95
Administrative and professional settings	2	1	2	14	30	42	8	3	182
Wholesalers, warehouses, manufacturing and agricultural settings	6	9	9	6	13	19	13	25	32
Retail settings with no further detail	6	4	6	9	14	21	22	18	2725
Banking and financial settings	2	0	17	23	15	34	4	6	223
Pharmacies and chemists	1	1	1	12	20	37	22	6	277
Service stations	28	20	3	2	3	4	13	26	819
Licensed premises	9	5	3	7	3	7	28	38	58
Unspecified and other locations	8	4	6	5	16	12	20	30	172
Note: Percentages may not total 100 because of rounding. Source: Australian Institute of Criminology NARMP 2003 [computer file]

Property stolen in armed robbery

The property stolen from victims was not recorded for all cases (valid cases n=3,336). The following discussion is, at best, broadly indicative of possible trends in the types of property stolen during armed robberies that occurred in 2003. This is because:

• of the varying detail in property descriptions;
• of the fact that these data were available for less than half of all victims; and
• up to only five property types could be recorded (and so sixth and subsequent types of property could not be considered).

Thirty-eight per cent of victim cases contained valid stolen property information. Of these, 1,237 recorded the theft of multiple property items, resulting in a total of 6,413 valid stolen property types listed in the NARMP (see Table 8). Over one-quarter (27%) of the reported property taken was categorised as cash, followed by mobile phones and/or accessories (16%). Given that wallet/purses, and handbag/briefcases made up eight per cent of reported property, it is not surprising that the common contents of these - credit and debit cards, and identity documents - were reported taken in roughly similar proportions (9% and 12% respectively).

Table 8 : Types of property taken in armed robberies

Property	Number	%
Cash	1,716	27
Negotiable documents
Negotiable documents (no further detail)	6	<1
ATM card	60	1
Credit card	538	8
Cheque or money order	30	<1
Other neg doc (not classified elsewhere)	93	1
Total	727	11
Identity documents
Identity documents (no further detail)	2	<1
Drivers licence	237	4
Other licence	20	<1
Passport	14	<1
Other ID document (not classified elsewhere)	495	8
Total	768	12
Luggage
Wallet or purse	399	6
Handbag or briefcase	99	2
Other luggage (not classified elsewhere)	150	2
Total	648	10
Personal electrical equipment
Electrical equipment (no further detail)	1	<1
Mobile phone or accessory	1,055	16
Personal stereo or TV or accessory	77	1
Laptop or PC or accessory	61	1
Personal organiser	6	<1
Camera or still or accessory	61	1
Camera or video or accessory	37	1
Electronic game or accessory	17	<1
Telecommunications excluding mobiles or accessories	28	<1
Other electrical (not classified elsewhere)	17	<1
Total	1,360	21
Jewellery
Jewellery (no further detail)	1	<1
Ring	73	1
Earrings	22	<1
Chain	115	2
Watch	100	2
Other jewellery (not classified elsewhere)	59	1
Total	370	6
Alcohol and drugs
Alcohol	11	<1
Tobacco	125	2
Drugs (illicit)	12	<1
Pharmaceuticals	8	<1
Prescription pad or form	3	<1
Other alcohol or drugs (not classified elsewhere)	5	<1
Total	164	3
Weapons
Handgun	61	1
Knife	9	<1
Syringe	1	<1
Total	71	1
Personal items not classified elsewhere
Personal items (no further detail)	17	<1
Other property (no further detail)	73	1
Clothing	65	1
Footwear	19	<1
Sports equipment	17	<1
Lighter or smoking accessory	6	<1
Book or publication or stationery	14	<1
Spectacles or binoculars	16	<1
Keyring or personal keys	90	1
Other personal items (not classified elsewhere)	21	<1
Total	338	5
Conveyance
Vehicle (no further detail)	7	<1
Bicycle	16	<1
Fuel	1	<1
Conveyance accessories	32	<1
Total	56	1
Other property not classified elsewhere
Cash register	20	<1
Safe or cash box	48	1
Foodstuffs or groceries	14	<1
Cosmetic, toiletry or perfume	9	<1
Furniture or homewares	6	<1
Tools or machinery or hardware	11	<1
Keys	7	<1
Documents (not classified elsewhere)	38	1
Retail goods (no further detail)	1	<1
Collectors items (not classified elsewhere)	11	<1
Medical equip excluding syringe	1	<1
Office equip excluding PCs or telecommunications	1	<1
Vending machine excluding poker machine	14	<1
Musical instrument	4	<1
Other property (not classified elsewhere)	10	<1
Total	195	3
Total	6,413	100
Note: Multiple property types were listed in some cases, therefore number refers to total property types, not total victim cases. Percentages may not total 100 because of rounding. Source: Australian Institute of Criminology NARMP 2003 [computer file]

The type of property stolen in robberies was not uniform across all locations (see Table 9). Cash constituted the overwhelming majority of the property taken in financial institutions and in licensed premises (70% and 73% respectively). Personal electrical equipment (mobile phones, stereos, electronic diaries, cameras, etc) accounted for between one-quarter and one-third of the property taken from what could be classified as muggings (robberies on the street and footpath, 28%; in transport related settings, 29%; and in open spaces, 26%). Personal electrical equipment made up a similar proportion of the property stolen in residential armed robberies (32%).

Table 9 : Property type taken in armed robberies, as percentage of location

Location	Cash	Negotiable documents	Identity documents	Luggage	Pers. electric equipment	Jewellery	Alcohol and drugs	Weapons	Pers. items NEC	Conveyance	Other prop. NEC	Number
Residential setting	20	9	13	8	32	5	1	1	3	1	7	290
Recreational settings	27	5	12	11	22	4	1	3	9	0	6	561
Transport-related settings	19	14	16	11	29	2	1	1	4	1	3	519
Open spaces excluding street and footpath	11	8	13	11	26	4	0	1	11	12	2	245
Street and footpath	18	10	14	15	28	5	1	0	5	1	3	2,077
Education, health, religious, justice and other community settings	11	18	27	7	11	9	2	14	2	0	0	56
Administrative and professional settings	25	19	18	2	21	4	0	8	1	0	2	191
Wholesalers, warehouses, manufacturing and agricultural settings	28	8	36	0	28	0	0	0	0	0	0	25
Retail settings with no further detail	34	15	10	7	13	9	3	0	2	0	6	1,684
Banking and financial settings	70	4	1	5	7	2	0	10	0	0	0	81
Pharmacies and chemists	38	25	7	8	3	2	11	0	1	0	5	122
Service stations	62	5	2	1	6	1	16	1	2	0	6	367
Licensed premises	73	0	0	5	5	0	14	0	0	0	5	22
Unspecified and other locations	17	12	9	8	24	20	0	5	3	1	1	173
Total %	27	11	12	10	21	5	3	1	4	1	4	6,413
Note: 'NEC' refers to items to 'not elsewhere classified'. Multiple property types were listed in some cases, therefore the total refers to total property types, not total number of victim cases. Percentages may not total 100 because of rounding. Source: Australian Institute of Criminology NARMP 2003 [computer file]

In general, the other types of property taken were consistent with the type of stock and items one would expect in the various locations. For instance, drugs and alcohol (presumably pharmaceuticals) in chemists (11%), in service stations (tobacco; 16%); and in licensed premises (alcohol and tobacco; 14%). A relatively high proportion of property taken in pharmacies was in the category 'negotiable documents' (25%), in other words credit and debit card related items. Both negotiable documents and identity documents account for a high proportion of the items taken in office-type locations (administrative and professional settings [19% and 18% respectively]; and education, health, religious, justice and other community settings [18% and 27% respectively]). Examination of detailed data showed that the high proportion of 'weapons' taken from the latter setting category (i.e. 14% from education, health etc settings) was primarily because of syringes that were stolen from health locations.

The value of the property stolen was also provided in some cases, even if actual property descriptions had not been supplied. In some instances, the total value was calculated as the sum of the value of each property type, whereas a single value was provided for other cases.¹⁰ The average total value of property taken in robberies was $6,464 (n=4,886), although there was considerable variation, with values ranging from zero to $550,000 (see Figure 2). Because values tended to cluster around the lower end of the range, a more accurate measure of the mid-range value is the median, which was $590.

Higher reported average values were associated with firearm robberies (mean=$14,085; n=1,203) than with 'other weapons' ($8,461; n=1,092) or knives ($2,200; n=2,283). The smallest average takings were associated with syringe robberies ($783; n=180). This pattern was generally consistent across all locations, and the exceptions that emerged - such as 'other weapons' resulting in higher average total values than firearms in financial and administrative/professional settings, and pharmacies - may simply be the result of the small number of cases contributing to average values (see Table 10).

Table 10 : Average total value of property taken in armed robberies in different locations, by weapon type

Location	Firearm		Knife		Syringe		Other weapon		Total
	avg $	n	avg $	n	avg $	n	avg $	n	avg $	n
Residential setting	16,775	34	964	97	1,395	1	3,148	72	4,372	204
Recreational settings	21,269	244	6,795	169	1,326	8	6,290	73	13,781	494
Transport related settings	8,843	48	683	148	646	22	3,369	79	2,714	297
Open spaces excluding street and footpath	3,037	14	599	80	593	4	1,153	35	1,001	133
Street and footpath	3,997	143	1,204	615	748	43	1,815	341	1,719	1,142
Education, health, religious, justice and other community settings	27,364	8	420	13	-	0	367	3	9,395	24
Administrative and professional settings	11,724	71	3,043	32	-	0	13,501	23	9,844	126
Wholesalers, warehouses, manufacturing and agricultural settings	350	1	6,793	8	-	0	1,596	9	3,836	18
Retail settings with no further detail	8,823	434	2,683	675	888	80	7,694	291	5,372	1,480
Banking and financial settings	78,273	46	15,427	28	149	2	221,389	22	90,850	98
Pharmacies and chemists	1,199	40	629	92	509	9	2,028	20	938	161
Service stations	5,967	82	1,277	260	445	8	981	91	2,073	441
Licensed premises	16,052	10	1,288	9	-	0	182	3	7,848	22
Unspecified and other locations	34,512	28	1,568	57	240	3	2,087	30	9,483	118
Total	14,085	1,203	2,200	2,283	783	180	8,461	1,092	6,588	4,758
Source: Australian Institute of Criminology NARMP 2003 [computer file]

There are a number of possible reasons as to why firearm robberies might result in higher average losses. Locations with relatively high value property may employ more security. Offenders may need highly dangerous weapons in order to overcome that security - to instil fear and so ensure compliance from victims - but this ultimately results in larger gains for robbers. Another possibility is that the apparent link between firearms and higher average gains is in fact the product of offender planning: more professional offenders who engage in substantial pre-offence planning and also target properties with the potential for high takings may simply opt for such weapons. Regardless, data again suggest a subset of robberies that can be described by the high gain/dangerous weapon 'professional armed robbery' scenario.

Outcomes of armed robbery investigations

Information contributing to the NARMP is extracted from police systems on a bi-annual basis, thus cases occurring towards the close of a six-month period cannot, by definition, have been investigated for as long as those occurring at the start of that same period. Because investigative outcome variables are not updated after the initial date of extraction, outcome data should not be treated as final, but rather as indicative of the state of robbery investigations at the time of data extraction. The reliability of this variable is further weakened because the length of time that elapsed between the last day of a data collection period and the time when data were extracted was not the same for all jurisdictions nor for the two reporting periods contributing to the data set. When considering the most recent outcome for the 6,142 victims for which data were available:

• 30 per cent of victim cases recorded that investigations were finalised, with offender(s) proceeded against;
• five per cent had been finalised (with no offender proceedings); and
• the majority (65%) were not finalised.

These same proportions held across the four main weapon types¹¹ (Table 11), although relatively wider variation was seen amongst the different location types (Table 12). For instance, around half of all robbery victimisations¹² occurring in financial settings (46%) and in pharmacies (50%) were finalised with the offender(s) proceeded against. In contrast, only twenty two per cent of street robberies had been finalised. This may reflect differences in the types of robberies that occur in locations. For example, financial institution robberies are relatively few in number and may involve a set of offenders who 'specialise' in this high-gain target. Police investigators may therefore be more familiar with likely offenders and employ intelligence-based strategies when policing these incidents. Street robberies are greater in number, appear to be more opportunistic (at least in terms of weapon use) and seem less effectively targeted (in terms of lower gains to offenders), and so are perhaps perpetrated by a wide variety of less-frequent offenders, all of which make their detection less likely. Unlike street locations, financial settings typically also have active security measures (e.g. security cameras, staff trained to note offender characteristics) that can assist investigators, and this might too be a factor in the differential clearance rates.¹³

Table 11 : Most recent investigative outcome status of armed robberies, as percentage of first listed type of weapon

Weapon	Not finalised	Finalised, no offender proceeded against	Finalised, offender proceeded against	Number
Firearm	65	3	33	1,144
Knife	65	6	29	3,084
Syringe	66	7	27	272
Other weapon	65	5	30	1,458
Total	65	5	30	5,958
Note: Most recent investigative outcome status refers to latest reported outcome at the time of data extraction from jurisdictional recording systems: this variable has not been refreshed since initial data extraction. The length of time elapsed from date of robbery to date of data extraction varies depending on when the robbery occurred during the time period examined and when data was extracted, which varied between jurisdictions. The first listed weapon was used rather than all weapons because some incidents employed multiple weapons. Percentages may not total 100 because of rounding. Source: Australian Institute of Criminology NARMP 2003 [computer file]

Table 12 : Most recent investigative outcome status of armed robberies, as percentage of location

Location	Not finalised	Finalised, no offender proceeded against	Finalised, offender proceeded against	Number
Residential setting	49	13	39	372
Recreational settings	78	2	20	440
Transport related settings	63	7	30	517
Open spaces excluding street and footpath	80	4	16	120
Street and footpath	72	6	22	1,661
Education, health, religious, justice and other community settings	56	9	35	66
Administrative and professional settings	78	2	20	105
Wholesalers, warehouses, manufacturing and agricultural settings	75	13	13	24
Retail settings with no further detail	62	4	35	1,842
Banking and financial settings	51	3	46	122
Pharmacies and chemists	48	2	50	143
Service stations	60	3	37	553
Licensed premises	56	4	40	55
Unspecified and other locations	74	7	19	122
Number	3,999	313	1,830	6,142
Note. Most recent investigative outcome status refers to latest reported outcome at the time of data extraction from jurisdictional recording systems: this variable has not been refreshed since initial data extraction. The length of time elapsed from date of robbery to date of data extraction varies depending on when the robbery occurred during the time period examined and when data was extracted, which varied between jurisdictions. Percentages may not total 100 because of rounding. Source: Australian Institute of Criminology NARMP 2003 [computer file]

Armed robbery victims

Of the 8,865 victims detailed in the NARMP, 2,289 were categorised as organisations. It is quite probable that a number of these organisational victimisations were targeted in the same incidents as individual victims, but there are no identifying variables that would allow these cases to be linked. Therefore for the purposes of this report, the robbery of organisations and of individual persons is explored separately.¹⁴

Individual victims of robbery

As already noted, targets of armed robbers are probably selected in a fairly opportunistic way. Given this, are there any features that tend to characterise people victimised by armed robbers? Analyses of victim age and gender show that victims were predominantly male (71% of all cases for which age and gender data was available) and tended to be aged in their late teens and early adult years (15 to 34 years; 66% of men and 60% of women; see Table 13). Victims ranged in age from zero¹⁵ to eighty-eight years, with an average of thirty years (median=26 years; n=6,368).

Table 13 : Victim age group, as percentage of victim gender

Age group	Male	Female	Number
<15 yrs	4	1	215
15 to 17 yrs	12	8	699
18 to 19 yrs	11	8	649
20 to 24 yrs	21	19	1,305
25 to 29 yrs	13	14	859
30 to 34 yrs	9	11	590
35 to 39 yrs	6	9	447
40 to 44 yrs	6	9	415
45 to 49 yrs	6	8	396
50 to 54 yrs	4	6	310
55 to 59 yrs	3	4	228
60 to 64 yrs	2	2	122
65 yrs and over	2	2	112
Number	4,498	1,849	6,347
Note: Percentages may not total 100 because of rounding. Source: Australian Institute of Criminology NARMP 2003 [computer file]

A closer examination of data on gender as a function of (first listed) weapon use showed that the manner in which individuals were robbed differed for males and females. Around half of the robberies experienced by men and by women involved knives (51% and 49% respectively), although women were slightly more likely to be robbed with syringes (7%, versus 3% for males) and with firearms (24%, versus 20% for males). Proportions of robberies involving 'other weapons' also varied marginally with gender (females=21%; males=26%).

Weapon use was found to vary based on the gender and age of victims (Table 14). At least half of the robberies within each major class of weapon occurred in the eighteen to thirty-four years age grouping. This was true for both males and females victims. Despite male youths (i.e. those under 18 years of age) constituting a relatively small proportion of all male victims (16%), around one-fifth of both knife robberies and 'other weapon' robberies against male victims were perpetrated against youths. The equivalent proportion was less than twelve per cent for girls under eighteen years robbed with any weapon type, and males under eighteen robbed with syringes or with firearms. A disparity between men and women was also observed among the 35 to 59 year age group. Women so aged constituted forty per cent of all female firearm robbery victims, and around one-third of those were subject to knife, syringe and 'other weapon' robberies. Only thirty-three per cent of male victims of firearm robberies, around one-quarter of males subject to 'other weapon' attacks, and close to one-fifth of men robbed with syringes or with knives fell into this age bracket.

Table 14 : Weapons used against male and female victims of different ages, as percentage of weapon type by gender

Age group	Male				Female				Total number
	Firearm	Knife	Syringe	Other weapon	Firearm	Knife	Syringe	Other weapon
<15 yrs	1	6	1	4	0	1	3	2	209
15 to 17 yrs	5	14	11	15	6	8	9	9	671
18 to 19 yrs	6	12	13	13	6	10	5	8	628
20 to 24 yrs	21	22	27	20	19	20	13	18	1,268
25 to 29 yrs	14	13	16	12	14	14	16	14	835
30 to 34 yrs	13	8	7	8	11	9	18	12	575
35 to 39 yrs	8	6	6	6	8	10	10	7	429
40 to 44 yrs	8	5	3	6	9	8	8	10	402
45 to 49 yrs	8	4	5	6	9	8	5	8	385
50 to 54 yrs	5	4	6	5	9	5	5	5	301
55 to 59 yrs	4	3	3	3	5	4	3	5	221
60 to 64 yrs	3	2	0	1	3	2	1	0	119
65 yrs and over	4	1	1	1	1	1	4	2	109
Number	856	2,227	151	1,127	421	879	117	374	6,152
Note: Percentage based on only first listed weapon. Percentages may not total 100 because of rounding. Source: Australian Institute of Criminology NARMP 2003 [computer file]

These slight differences may be a function of the locations that victims frequent because of factors such as lifestyle: male youths may more often find themselves in those unguarded open spaces that are sites of muggings, whereas women in their middle adult years (by virtue of their professions) may be in those locations more likely to be the sites of firearm robberies, namely retail settings. An examination of victim age and gender as a function of location adds some support for this possibility: male youths were most often victimised on the street and in transport related settings (total of 65%), whereas forty-two per cent of women aged thirty-five to fifty-nine were victimised in generic retail locations (see Table 15).

Table 15 : Locations where males and females of different ages were victimised, as a percentage of gender by age group

Location	Male				Female				Number
	<18	18-34	35-59	60+	<18	18-34	35-59	60+
Unspecified and other locations	3	2	2	2	4	2	2	0	131
Residential setting	3	7	10	12	3	6	8	14	467
Recreational settings	8	9	11	17	5	8	7	5	562
Transport related settings	16	10	6	6	15	8	4	11	578
Open spaces excluding street and footpath	6	3	2	1	5	2	1	2	179
Street and footpath	49	40	24	15	27	24	13	20	2017
Education, health, religious, justice and other community settings	2	1	1	4	2	1	2	0	79
Administrative and professional settings	<1	1	3	4	1	2	5	6	114
Wholesalers, warehouses, manufacturing and agricultural settings	<1	<1	1	0	0	0	<1	0	20
Retail settings with no further detail	12	15	30	27	34	34	42	35	1532
Banking and financial settings	0	2	2	1	1	3	4	3	126
Pharmacies and chemists	<1	<1	2	7	2	5	7	6	156
Service stations	<1	9	7	5	1	3	3	0	365
Licensed premises	<1	1	<1	0	1	<1	<1	0	21
Number	750	2,437	1,143	168	164	966	653	66	6347
Note: Percentages may not total 100 because of rounding. Source: Australian Institute of Criminology NARMP 2003 [computer file]

Organisational victims of robbery

Organisations were threatened with firearms in a slightly higher proportion of cases than individual victims (29%, versus 21% for persons), and with knives and 'other weapons' slightly less often (46% and 21% respectively for organisations, versus 50% and 24%; see Table 16). There was also variation between organisations and individuals in the number of offenders involved in their victimisation. Somewhat unexpectedly, organisational victims were subject to a higher proportion of attacks by single offenders (68%) than were individuals (57%; Table 17), although the numerous limitations of offender data may lie behind this pattern rather than any genuine differences (see section below for the detail of offender data limitations).

Table 16 : Weapon use against organisational and individual victims, as percentage of victim type

Weapon	Organisation	Individual	Number
Firearm	29	21	1,974
Knife	46	50	4,213
Syringe	4	4	374
Other weapon	21	24	1,990
Number	2,227	6,324	8,551
Note: Percentages may not total 100 because of rounding. Source: Australian Institute of Criminology NARMP 2003 [computer file]

Table 17 : Number of offenders victimising organisational and individual victims, as percentage of victim type

No. of offenders	Organisation	Individual	Number
1	68	57	1,296
2	22	25	510
3	6	10	191
4	2	5	86
5	2	3	56
Total	684	1,455	2,139
Note: Percentages may not total 100 because of rounding. Source: Australian Institute of Criminology NARMP 2003 [computer file]

A greater proportion of organisational victims were robbed by offenders in older age groups, relative to individual victims: sixty-eight per cent of the reported armed robbers victimising organisations were aged between twenty and forty-nine years, in contrast to the fifty-two percent of those similarly aged robbing individual persons. Stated differently, nearly half (48%) of the reported offenders robbing individuals were aged under twenty years, whereas only thirty-one per cent of armed robbers victimising organisations were in their teens or younger (see Table 18). This could occur because:

• older offenders may have the necessary experience to effectively target the (presumably) more secure organisational victims;
• older offenders may have less fear of the consequences of capture associated with more secure (and therefore presumably harder to breech) premises because of a longer history of escaping detection. Alternatively, they may be less fearful because they may have experienced the justice consequences of armed robbery on multiple occasions, minimising any deterrent effect detection may have once held; or
• older offenders may be more 'professional', and so only target victims known to provide guaranteed and sizeable gains.

Table 18 : Age group of offenders victimising organisational and individual victims, as percentage of victim type

Offender age	Organisation	Individual	Number
<15 yrs	2	6	170
15 to 17 yrs	14	24	734
18 to 19 yrs	15	18	591
20 to 24 yrs	28	25	900
25 to 29 yrs	16	10	403
30 to 34 yrs	13	10	384
35 to 39 yrs	7	4	165
40 to 44 yrs	3	2	77
45 to 49 yrs	1	1	40
50 to 54 yrs	<1	<1	8
55 to 59 yrs	<1	<1	5
60 to 64 yrs	<1	<1	2
65 yrs and over	0	0	0
Number	1,010	2,469	3,479
Note: Total differs to other analyses because some victim cases involved multiple offenders: if multiple offenders were listed, the associated victim was counted once for each offender. Percentages may not total 100 because of rounding. Source: Australian Institute of Criminology NARMP 2003 [computer file]

Victim resistance to offenders and victim injury

Victims may meet offenders' demands with complete compliance or with full or partial resistance. Although the NARMP can record information regarding victim resistance, it is inconsistently recorded by police and difficult to extract from police recording systems. Only sixty-three cases contained information regarding resistance, and of these, the majority (59%) were annotated as 'no resistance noted'. Fourteen per cent involved physical resistance, eleven per cent escape by the victim, and the remainder, verbal resistance to offender demands. The minimal data meant no further analyses involving resistance were undertaken.

Similar recording and extraction issues relate to the variable of victim injury. Of the 1,135 cases containing a non-missing value, over half were coded as 'no injury noted' (50%) or not applicable (2%). One-third were annotated as involving a minor injury, three per cent involving a serious injury, and twelve per cent resulted in emotional trauma.

Although injury data were only available for around one-eighth of victims, the relative proportions of knives, firearms, syringes and other weapons used against this subset of victims was generally similar to that seen across all cases in the NARMP (Table 19). For instance, knives were involved in the highest percentage of cases within each injury category.¹⁶ However, the relative proportions of general weapon use were not mirrored within each of the injury categories. For example, firearms made up about one-fifth of all weapons used in the NARMP, and were also the weapons used against twenty per cent of victimisations resulting in emotional trauma, but firearms were involved in only eight per cent of the cases associated with serious injury. In contrast, 'other weapons' were used in forty-two per cent of cases in which serious injury were recorded, but they constituted only twenty-four cent of all weapons in all NARMP cases. Injury data are quite unreliable, and apparent patterns may not be borne out when better quality data are accumulated. However if they are genuine effects, differences may be due to more lethal weapons exacting better victim compliance; less threatening weapons may necessitate additional measures such as actual violence in order to elicit the same compliance. Alternatively, less experienced or less professional offenders, who are more prone to respond to victims with actual violence, may use more opportunistic weapons.

Table 19 : Weapon type, as percentage of injury

Injury	Firearm	Knife	Syringe	Other weapon	Number
No injury noted	14	62	7	17	560
Minor injury	14	56	3	26	373
Serious injury	8	50	0	42	38
Emotional trauma	19	63	4	14	129
Total %	15	60	5	20	1100
Note: Percentage based on first listed weapon type only. Percentages may not total 100 because of rounding. Source: Australian Institute of Criminology NARMP 2003 [computer file]

Armed robbery offenders

Police information systems are not all equally capable of extracting the details of multiple offenders when compiling victim files, and even when able, the NARMP records details of only up to five offenders per victim. Logically, offender information detailing more than simply the gender of offenders can only be reported if offenders have been apprehended. These factors combine to mean that offender information is neither completely reliable nor truly representative of all armed robbers in Australia. Mindful of these factors, valid data describing the gender of 3,540 reported offenders were recorded in the NARMP. These offenders were linked to 2,156 victim cases, of which the average number of offenders per victimisation was 1.6 (median=1). The reported number of offenders ranged between one and five.

An examination of the number of offenders associated with victimisations in different locations showed that of the cases where the number of offenders was available, single offenders were involved in at least half of the robberies in virtually all locations (see Table 20). In general, offender pairs were reported in between one-fifth and one-quarter of cases, although there is the suggestion¹⁷ that groups of offenders (i.e. two or more) did not equally target all locations. Multiple offenders were responsible for around one-quarter of robberies in pharmacies, and three in ten in service stations and licensed premises, and over one-third of those in unspecified retail settings. In a very different pattern, more than one offender was linked to around half of the robberies in recreational settings (51%) and on the street (49%), and in transport related locations (43%), and financial settings (44%).

Table 20 : Number of offenders, as percentage of location

Location	Number of offenders					Number
	1	2	3	4	5
Residential setting	62	21	13	3	1	165
Recreational settings	49	20	9	17	5	108
Transport related settings	57	27	14	0	1	176
Open spaces excluding street and footpath	55	27	14	5	0	22
Street and footpath	51	29	11	6	2	424
Education, health, religious, justice and other community settings	58	31	8	0	3	36
Administrative and professional settings	71	29	0	0	0	28
Wholesalers, warehouses, manufacturing and agricultural settings	42	8	0	0	50	12
Retail settings with no further detail	64	23	8	2	3	752
Banking and financial settings	56	23	0	19	2	64
Pharmacies and chemists	77	9	11	3	0	75
Service stations	69	22	5	1	3	242
Licensed premises	70	17	4	9	0	23
Unspecified and other locations	62	28	3	0	7	29
Total %	60	24	9	4	3	2,156
Note: Percentages may not total 100 because of rounding. Source: Australian Institute of Criminology NARMP 2003 [computer file]

It seems unlikely that the same types of groups would target both financial settings and perpetrate muggings in open public locations, because these settings have very different levels of security. The former is perhaps the most secure of all sites, but the latter is usually unguarded or otherwise unprotected. Unfortunately, the number of cases is too small to permit a meaningful exploration of other ways these multiple offender cases might differ with the various locations (e.g. offender disguises).

The relative proportions of weapons used varied for offenders acting alone and in groups of different sizes (Table 21). Robberies by single offenders and by offender pairs and trios involved knives in around half of all cases (53%, 49% and 53% respectively), with firearms used in around one-fifth of cases (21% for solo robbers, 18% for pairs and 21% for trios). Larger groups of offenders showed slightly different weapon usage: groups of five offenders were associated with slightly more firearm victimisations (27%) and fewer knives (39%), with groups of four robbers associated with the largest proportion of firearms (45%) and the smallest of knives (31%). Syringes were not used when more than three offenders were involved in an incident. Varying patterns of weapon use may be one way to differentiate the larger offender groups targeting seemingly highly disparate locations (street versus financial settings), and as more data accumulates in the NARMP, this possibility can be examined.

Table 21 : Weapon type by single and multiple offenders, as percentage of number of offenders

No. of offenders	Firearm	Knife	Syringe	Other weapon	Number
1	21	53	5	21	1,269
2	18	49	4	29	503
3	21	53	0	26	189
4	45	31	0	25	85
5	27	39	0	34	56
Total %	21	51	4	24	2,102
Note: Percentage based on first listed weapon type only. Percentages may not total 100 because of rounding. Source: Australian Institute of Criminology NARMP 2003 [computer file]

As with armed robbery victims, the overwhelming majority of offenders were male (91%; valid n=3,212). The majority of the male offenders (where information on age was available) were in their teens and early to mid-adult years (15 to 34 years, 87%). Although females constituted a much smaller proportion of reported offenders (9%), the majority of these were also aged from 15 to 34 years (79%; see Table 22). In cases where at least two offenders were involved, the majority of offender pairs (80% of 851 cases) were male. Only three per cent involved two females, with the remainder made up of male-female pairs.

Table 22 : Offender age group, as percentage of offender gender

Age group	Male	Female	Unknown	Total %
<15 yrs	5	8	8	5
15 to 17 yrs	22	18	8	21
18 to 19 yrs	17	13	0	17
20 to 24 yrs	26	21	8	26
25 to 29 yrs	11	16	8	12
30 to 34 yrs	11	11	25	11
35 to 39 yrs	4	7	0	5
40 to 44 yrs	2	2	8	2
45 to 49 yrs	1	2	33	1
50 to 54 yrs	<1	0	0	<1
55 to 59 yrs	<1	0	0	<1
60 to 64 yrs	<1	0	0	<1
65 yrs and over	0	0	0	0
Number	3,179	314	12	3,505
Note: Percentages may not total 100 because of rounding. Source: Australian Institute of Criminology NARMP 2003 [computer file]

Characteristics of offenders and weapon use

The victim-based organisation of the NARMP means that the weapons used by each and every offender cannot be recorded. Rather, weapons used against a victim, and the offender(s) who robbed that victim are recorded, but individual weapons are not linked to individual offenders. In order to gain some picture of any differences in weapon use as a function of the offender gender and age, the first listed offender's age and gender was cross-tabulated with the first listed weapon. Results show that knives generally made up the largest proportion of weapons associated with both male and female offenders (Table 23). A slightly different pattern observed among males aged 30 to 44 years may reflect a sort of 'professionalisation' - an increase in the planning that robbers engage in - where they use proportionally more harder to obtain weapons (such as firearms) relative to their younger counterparts. Apparently anomalous findings among men in middle and later adulthood (i.e. 45 years and above) and women of all ages are likely to be as a result of the very small number of cases contributing to these subgroups.

Table 23 : Weapons used by males and female offenders of different ages, as percentage of age group for gender

Age group	Male				Female				Total no.
	Firearm	Knife	Syringe	Other weapon	Firearm	Knife	Syringe	Other weapon
<15 yrs	3	61	4	32	0	71	14	14	76
15 to 17 yrs	11	58	2	29	0	74	0	26	324
18 to 19 yrs	18	50	2	30	37	42	0	21	304
20 to 24 yrs	19	54	4	23	18	36	24	21	529
25 to 29 yrs	16	59	5	20	44	33	0	22	290
30 to 34 yrs	40	36	5	19	4	58	21	17	315
35 to 39 yrs	32	43	3	21	10	50	20	20	131
40 to 44 yrs	38	35	0	27	60	40	0	0	60
45 to 49 yrs	17	71	4	8	0	0	14	86	31
50 to 54 yrs	43	57	0	0	0	0	0	0	7
55 to 59 yrs	80	20	0	0	0	0	0	0	5
60 to 64 yrs	0	0	0	100	0	0	0	0	1
65 yrs and over	0	1	2	3	0	0	0	0	0
Total % for gender	21	51	4	24	19	47	11	23	2,073
Note: Percentage based on first listed weapon type only, and calculated for only first listed offender. Percentages may not total 100 because of rounding. Source: Australian Institute of Criminology NARMP 2003 [computer file]

A similar examination of location as a function of the first listed offender also hints at a sort of professionalisation of male robbers as they grow older (Table 24). Nearly half (49%) of all armed robberies by youths (i.e. under 18 years) occurred in transport related settings and on the street (i.e. muggings), whereas only around three in ten (29%) took place in retail settings of any kind. Amongst eighteen and nineteen year old males, the proportions shifted, with four in ten robberies (41%) occurring in any retail setting, and only forty-three per cent in street and transport locations. For offenders aged between twenty and thirty-nine, one-fifth (20%) of robberies were street or transport muggings, whereas sixty-five per cent took place in the retail settings. Among these older offenders, at least one in eight (13%) armed robberies took place in a service station. Of course given that this data set is only a snapshot of a subset of all offenders it impossible to unequivocally state if these patterns reflect the 'maturation' of individuals, or if they are simply a cohort effect, where unknown generational factors cause offenders of different ages to select different locations. At present, an offender identifier is not available to allow individuals to be tracked over time; however this possibility could be explored in the near future if appropriate offender information were to be incorporated into the NARMP.

Table 24 : Locations robbed by male offenders of different ages, as percentage of age group

Location	<15	15-17	18-19	20-24	25-29	30-34	35-39	40-44	45-49	50+	No.
Residential setting	1	6	6	6	9	8	9	14	13	8	141
Recreational settings	13	9	5	2	2	4	1	11	13	0	93
Transport related settings	20	14	15	6	5	6	2	0	4	0	163
Open spaces excluding street and footpath	4	2	2	<1	0	1	0	4	0	0	20
Street and footpath	39	33	28	17	18	8	13	7	13	8	388
Education, health, religious, justice, other community settings	1	2	2	<1	<1	3	0	4	0	0	25
Administrative and professional settings	0	0	0	0	1	6	0	4	0	0	24
Wholesalers, warehouses, manufacturing, agricultural settings	0	1	2	<1	<1	<1	0	0	0	0	12
Retail settings with no further detail	16	22	29	42	42	41	42	39	38	31	697
Banking and financial settings	0	1	1	4	3	4	9	5	0	15	63
Pharmacies and chemists	0	1	0	5	6	6	5	2	0	0	68
Service stations	6	7	9	15	12	10	16	7	21	31	220
Licensed premises	0	1	2	1	1	1	1	2	0	8	20
Unspecified and other locations	0	2	1	2	<1	1	2	4	0	0	25
Number	70	307	287	507	269	298	127	57	24	13	1,959
Note: Offender categories capturing those aged 50 years and over have been collapsed into a single category because of the very small number of cases. Percentages may not total 100 because of rounding. Source: Australian Institute of Criminology NARMP 2003 [computer file]

Disguise use by offenders

A minimal amount of offender disguise information was available in the NARMP: only 129 victim cases contained any sort of disguise information, providing a total of 175 records regarding disguise (due to multiple disguise elements or multiple offenders). Disguise information is not a mandatory reporting field in most jurisdictions, and even when this information is recorded in police systems, it is difficult to extract in a victim-based format. This is unfortunate, because disguise can be (an albeit limited) proxy for the level of offender planning (see Borzycki 2003), as well as providing useful information about robbery modi operandi. Given the paucity of disguise data, comment is limited to noting that:

• forty per cent of these records referred to a mask;
• one in twenty (5%) to an improvised face covering;
• sixteen per cent to other accessories or simply the presence of an unspecified disguise; and
• a disguise was not seen or noted, or not applicable in the remainder of cases.

The relationship between offender and victim

As already mentioned, armed robbery can be thought of as more impersonal than other violent offences such as murder or assault, because often no relationship exists between offender and victim prior to the incident. When considering only robberies against individual persons (i.e. excluding organisations), the parties were unknown to each other in around nine out of ten (91%) of the 1,779 specified offender-victim relationships.¹⁸

An exception to this pattern emerged with robberies in residential locations: of the 179 relationships between individual victims and offenders linked to this location category, fifty-one per cent reported that parties were unknown, but three per cent recorded that offenders were related to victims, and the remainder (45%) were known non-family members. Residential armed robbery appears unlike that which occurs in other settings, because the level of anonymity typically associated with the offence seems greatly reduced. This has implications for crime reduction strategies targeting this location, because traditional target hardening is neither possible nor likely to deter the offender committed to obtaining property from a specific residential location.

---

• ⁵ The technical appendix to this report contains a summary of the variables supplied, the valid cases for each, and technical points pertaining to cases (Table 25). A separate table details the specific data values that were collapsed in order to create the categories employed in analyses (Table 26).
• ⁶ The NARMP does not contain identifier information that would allow victims involved in the same incident to be linked together.
• ⁷ Detailed location descriptions were not available for the majority of cases, thus findings regarding specific locations outside standard ABS coding do not necessarily reflect all robbery victimisations that occurred in those locations.
• ⁸ Includes cases where weapon information included a valid value that at least specified a type of weapon or type of location (i.e. not simply 'unknown'; valid n=9204)
• ⁹ Season was calculated on the basis of calendar month; valid n=8,852.
• ¹⁰ As with the property type variables, the reliability of conclusions drawn regarding this variable are severely limited by missing data (which limits the number cases contributing to analyses and also because of unknown yet possibly systematic reasons why data are missing). Additionally, its validity is questionable because police officers in the jurisdictions supplying this variable are usually not required to obtain confirmation of initially entered values.
• ¹¹ Percentage based on first listed weapon type to overcome any double counting because of the multiple weapon types used in some cases.
• ¹² For which most recent investigative outcome data were also available.
• ¹³ Other research has suggested that offenders employing firearms may have their offending detected less often over the course of their careers, possibly because of their professionalism (see Mouzos & Borzycki 2003, p. 35). Data relating to offender criminal histories are not available in the NARMP.
• ¹⁴ If an organisation were subject to armed robbery, and individual(s) were also robbed of personal property in the same incident, both organisation and individual(s) would be counted as separate victims
• ¹⁵ Only four individual victims had a recorded age of zero.
• ¹⁶ Proportions were not identical to the national weapon use pattern, which may be a function of the relatively small number of cases with valid injury d