The Facts and Figures that Describe Australia's Crime Patterns and Trends

(This is an abbreviated version of my Chapter 1. Trends in Crime and Criminal Justice, in The Australian Criminal Justice System - The Mid-1990s, Eds. Chappell, D, and Wilson, P., 1994. The data are a little old now, and I haven't had time recently to update them; nevertheless they tell an interesting story which is no less relevant as time passes…….)

Satyanshu Mukherjee and Dianne Dagger (Australian Institute of Criminology) have assembled official statistics of crimes reported to police from 1973/4 to the most recent, and the story told in these figures is one of apparently ever-increasing levels of crime. By contrast, the national crime victimisation surveys (Australian Bureau of Statistics) show much more stable trends in crime, leading to the conclusion that there have been considerable increases in reporting in this period.

Total property crimes reported to police (including burglary, motor vehicle theft, fraud and stealing) increased from 385,453 in 1973/4 to 1,168,423 in 1990/1 (an increase of 203%) before falling in 1991/2 to 1024569, while total violent offences rose from a mere 7,056 to 36,909 in 1991/2 - an increase of 423%. Expressed as an annual rate per 100,000 population, property offending went from 2834.4 crimes reported per 100,000 to 6563.8 in this period; violence increased from 51.9 to a startling 213.4; even adjusted for population change, then, these are increases of 132% and 311% respectively. Table 1 shows the components of these trends: - of particular concern are the trends in reported sexual assaults (rape rates up 426%) and other serious assaults (up 452%). In addition, reported drug offences were up 612% between 1974/5 and 1991/2.

Table 1. Numbers of Crimes Reported to Police and Rate Per 100 000 Population
Australia 1973-74 to 1991-92

Offence Year	Homicide		Serious Assault		Rape		Robbery		Burglary (Dwelling)
Offence Year	Number	Rate	Number	Rate	Number	Rate	Number	Rate	Number	Rate
1973-74	294	2.16	2822	20.75	765	5.63	3175	23.35	55090	445.61
1974-75	265	1.92	2995	21.75	723	5.25	2921	21.21	55312	441.81
1975-76	286	2.06	3374	24.25	828	5.95	2855	20.52	57950	457.96
1976-77	317	2.25	3513	24.96	864	6.14	2970	21.10	59884	467.94
1977-78	306	2.15	4258	29.86	867	6.08	3472	24.35	69626	536.96
1978-79	285	1.98	4796	33.25	911	6.32	3509	24.33	84616	586.72
1979-80	323	2.21	5424	37.11	1083	7.41	4264	29.17	97226	665.21
1980-81	265	1.78	6250	41.88	1318	8.83	4601	30.83	112667	754.98
1981-82	305	2.01	6925	45.61	1552	10.22	5126	33.76	123924	823.20
1982-83	318	2.07	7782	50.55	1674	10.87	6606	42.91	148908	975.96
1983-84	332	2.13	8335	53.5	1846	11.85	6624	42.52	163655	1060.13
1984-85	308	1.95	9144	57.92	1886	11.95	6670	42.25	163939	1048.22
1985-86	344	2.15	11095	69.26	2104	13.13	6585	41.11	153057	955.51
1986-87	324	1.99	12698	78.15	2278	14.02	7967	49.03	169872	1045.44
1987-88	379	2.29	14768	89.33	2561	15.49	7985	48.3	174145	1053.39
1988-89	350	2.08	16910	100.61	2822	16.79	8463	50.35	180131	1081.83
1989-90	329	1.93	17963	105.13	2806	16.42	9015	52.76	163357	1068.04
1990-91	356	2.05	19668	113.45	3532	22.54	11135	64.23	191302	1113.70
1991-92	356	2.03	20075	114.52	4698	29.61	11780	67.20	194319	1119.33

Offence Year	Burglary (Other)		Motor Vehicle Theft		Larceny		Fraud		Drug Offence
Offence Year	Number	Rate	Number	Rate	Number	Rate	Number	Rate	Number	Rate
1973-74	64636	434.7	50943	374.60	182994	1478.46	31790	233.77	n/a	n/a
1974-75	67370	449.0	49673	360.70	182013	1452.00	30622	222.36	9734	70.7
1975-76	68249	448.9	51358	369.07	182904	1446.54	35349	254.03	16389	117.8
1976-77	74506	486.9	56772	403.38	220342	1577.42	31767	225.71	19596	139.2
1977-78	82789	531.9	64266	450.70	244870	1717.30	37419	262.42	21471	150.6
1978-79	78465	544.1	68726	476.54	273404	1895.75	42461	294.42	19713	136.7
1979-80	86045	588.7	70096	479.59	305343	2089.12	48959	334.97	23303	159.4
1980-81	90416	605.9	76455	512.32	305436	2046.71	42732	286.34	27642	185.2
1981-82	102164	665.8	86955	572.67	316497	2084.38	50182	330.49	34903	229.9
1982-83	112323	721.1	95796	622.32	296980	2117.59	57289	372.16	41866	272.0
1983-84	112627	713.2	99146	636.39	355718	2283.26	64106	411.48	51637	331.4
1984-85	107771	652.7	103164	653.42	377522	2391.15	68024	430.85	64717	409.9
1985-86	114593	715.4	752.70	134218	2550.7	452829	504.91	92284	390.2	62333
1986-87	135564	834.3	134218	826.02	452829	2786.84	92284	567.94	62333	383.6
1987-88	138337	836.8	123176	745.08	454706	2750.47	122779	742.68	48437	301.2
1988-89	149576	879.9	127194	756.80	454080	2701.77	123439	734.46	62696	379.3
1989-90	178367	932.0	136220	797.25	474154	2775.07	99180	580.47	59087	355.3
1990-91	178863	1041.3	149758	863.86	526710	3038.26	121790	702.53	71779	425.1
1991-92	167404	944.2	140467	801.34	523251	2985.06	125128	713.83	86470	503.5

Source: Compiled by Dianne Dagger from Police Annual Statistical Reports.
Notes: Where a particular item of data is not available for a given jurisdiction for a specific year, the previous year's data published by that jurisdiction has been used to calculate the national figure. These gaps in the data do not affect the general trends indicated by the data.
Homicide = Murder; Manslaughter (not by driving); Infanticide. (Homicide figures do not include attempts unless otherwise stated.)
South Australia 1973-74 to 1977-78 includes attempted murder; Northern Territory 1976-77 includes manslaughter by driving
Rape: Northern Territory 1988-89 and 1989-90 = Assault Sexual; Western Australia 1988-89 and 1989-90 = Assault Sexual Aggravated and includes Attempt.
1989-90 data supplied by the Crime Research Centre, Nedlands, W.A.

These apparent increases in reported crime rates are not consistent across jurisdictions. Overall risks of crime, as measured by the rates per 100,000 population, increased by around 40 per cent in Queensland, but by around 100 per cent in Western Australia.

Significantly, however, the national statistics for homicide remained remarkably steady within a range between 1.62 per 100,000 and 2.40 per 100,000, strongly refuting the very commonly held view that such incidents are on the increase. There is little evidence to suggest that rates of reported homicides do not reflect trends in the actual incidence of homicides, but even to the untrained statistical observer, it must be difficult to reconcile apparent increases in non-lethal violence (i.e. assaults and sexual assaults) of over 400 per cent with the fact that homicide itself has not increased at all. The more sophisticated observer must conclude that there is something wrong with the data on non-lethal violence. Unfortunately, there is considerable reluctance, particularly in some sections of the media and amongst some political groupings, to admit that crimes of violence have not been increasing rapidly.

But what do crime victims surveys have to say on this?

While clearly they are measuring different things, (victim surveys invariably bring in all sorts of trivia which respondents say were not reported to police precisely because of their lack of consequence) one would think they should tell roughly the same story as police statistics about relative trends in violence and property crime over time.

But this turns out not to be the case. One can compare the results of the A.B.S.' 1975 and 1983 National Crime Victims Surveys with those of their 1993 survey. Table 2 presents the figures, and shows that risks of break and enter changed from 25 per 1000 households, in 1975, to 61 in 1983 and 68 in 1993. These estimates include attempted burglaries. Rates of motor vehicle theft apparently remained at around 17 per 1000 households through the whole eighteen year period. Risks of robberies appear to have reduced between 1975 and 1983 then increased sharply from 1983 to 1993. Serious assaults in 1993 were lower than either of the two previous surveys and sexual assaults remained virtually constant at 0.5-0.6% (of females aged 18 years and over only). These figures suggest increases in these eighteen years of 51% for break, enter and steal, and no increase at all for motor vehicle theft - considerably different from the police figures of 153% and 122% increases respectively between 1974/5 and 1991/2 (the latest figures currently available). Even more striking, however, are the differences between survey and police statistics on violent offences. Police figures suggest a 219% increase in robbery, while surveys suggest only 33%; for other serious assault the police have an increase of 423% yet the surveys suggest a decrease of 9%. Survey figures on sexual violence are only available on a comparable basis for 1983 and 1993 (and must be treated with due caution as crime victimisation surveys are acknowledged even by their proponents as being problematic in this area because of the sensitive nature of the issue), but suggest virtually no change in incidence where police statistics suggest a 155% increase.

Table 2. Crimes reported to police and the corresponding crime victim survey results.
Crimes per 1000 persons/households, Australia, 1974/5 to 1992/3

Police Statistics	1974/5	1982/3	1991/2	Surveys	1975	1983	1993
Break, Enter & Steal	4.42	9.76	11.19	Break, Enter & Steal [a]	45	61	68
Motor Vehicle Theft	3.61	6.22	8.01	Motor Vehicle Theft [b]	17.6	17.6	17
Robbery	0.21	0.43	0.67	Robbery [c]	9	6	12
Sexual Assault	0.05	0.11	0.28	Sexual Assault	n/a	5	6
Other Serious Assault	0.22	0.51	1.15	Other Serious Assault [d]	27.5	34	25

Notes:
a. The survey results for break, enter and steal and for motor vehicle thefts are rates per 1000 households. The figure of 34 per 1000 households presented in the 1975 report excluded thefts from garages and sheds. However, the report of the 1983 Survey suggests that this figure should be inflated by at least a third to attain comparability with the 1983 result, which included both garages and sheds.
b. The 1975 and 1983 survey results presented motor vehicle thefts as rates per 1000 persons aged 15 and over. These figures have been adjusted to rates per 1000 households by multiplying by the average household size which was 2.2 persons in 1981.
c. The figure presented in the 1975 report was 2 and excluded incidents where no property was taken. However, the report of the 1983 Survey suggests that this figure should be inflated by a factor of `almost five' to attain comparability with the 1983 result, which included such incidents.
d. The figure presented in the 1975 report was 10 and specifically excluded threats of violence. However, the report of the 1983 Survey suggests that this figure should be inflated by a factor of at least 2.5 to attain comparability with the 1983 result, which included threats.

The principal reason for these differences is, it appears, a significant increase in the proportion of crimes which are reported to police. In the U.K., where four national crime victims surveys have been conducted since 1981, the comparisons made between survey results and police statistics make it very clear that this is the key explanation. In 1981 police in England and Wales recorded 3 million `notifiable offences' (6 per 100 population) and in 1991 they recorded 5.3 million (10 per 100 population). `Notifiable offences' includes burglary, theft, vandalism and violent offences. The three British Crime Surveys since then have shown increasing divergence between trends revealed by police records and those suggested by survey results. Recorded crime figures in England and Wales, as appears to be the case in Australia, nearly doubled between 1981 and 1991 while BCS results suggested a rise of only 50%.

Although such regular comparisons have only recently been possible in Australia, already similar results are being found. There are a number of reasons why this should not surprise us. Firstly, in response to public demand, the police have increasingly adopted policies which encourage reporting of crime, particularly, for example, in the areas of domestic violence and sexual assaults. Secondly, major lifestyle changes over the last two or three decades, such as the mass ownership of motor vehicles, and the combined effect of the increasing proportion of women in the workforce and the growth of dormitory suburbs, which now leaves residential properties unattended for much of the working day, have led to increases in the use of insurance to protect against property losses due to crime. One of the main reasons quoted by people for not reporting crimes to police is that they "know the police couldn't do anything about it", but as insurance companies usually require policy-holders to report thefts to police before they will pay on the losses, these apparently `hopeless' cases are now being reported anyway.

How does Australia fare in International Comparisons of Crime Victimisation?

What do the International Crime Victims Surveys tell us about crime in Australia in the late 1980s and early 1990s, compared to other countries? Table 3 presents a summary of the findings from the two surveys, in the form of aggregated data on estimated percentages of the population victimised in the previous twelve months, for 21 countries and nine different offence types.

At first sight, one could conclude from this Table that Australia, indeed, has a very serious crime problem. In fact, only bicycle owners in Australia would feel comfortable with these figures. Car theft is virtually double the average rate for the 21 countries, and, in most of the other offences including burglary and violence, Australian risks are at least fifty per cent higher than average. The perennial journalistic cliche that `crime in Australia may reach the levels of those in the U.S.A.' seems not only to have been achieved but to have been exceeded in most offence types. However, when overall rates of victimisation are plotted against levels of urbanisation, Australian victimisation levels, with an urbanisation rate of 63.5%, are clearly somewhat lower than one might expect from the line of best fit. The other countries at the same overall level of victimisation - the Netherlands, U.S.A. and Canada - have no such `excuse' for their crime rates.

Table 3. International Comparison of Victimisation Rates:
Percentages Victimised Once or More per Annum in 1988 and 1991.

	Car Theft	Theft From Car	Car Damage	Bike Theft	Burglary	Theft of Personal Property	Robbery	Sexual Assault	Other Assault
Australia	2.71	6.73	9.10	2.02	4.03	5.72	1.08	1.87	2.77
England	2.75	7.11	8.71	2.03	2.53	3.65	0.90	0.32	1.11
Scotland	0.81	5.35	6.46	1.02	2.03	2.63	0.54	0.79	1.08
N. Ireland	1.62	4.00	4.43	1.56	1.11	2.16	0.49	0.40	1.13
Netherland	0.41	6.02	8.87	8.75	2.16	4.53	0.93	0.88	2.04
W. Germany	0.43	4.68	8.70	3.32	1.29	3.95	0.83	1.74	1.86
Switzerland	n/a	1.86	4.09	3.24	0.97	4.53	0.50	n/a	0.70
Belgium	0.95	3.30	6.36	2.75	2.20	3.56	0.98	0.62	0.70
France	2.35	6.01	6.43	1.37	2.37	3.56	0.39	0.60	1.46
Norway	1.07	2.80	4.56	2.76	0.75	3.15	0.49	0.55	1.60
Finland	0.55	2.83	4.79	4.00	0.62	3.77	0.86	0.46	1.73
Spain	1.29	9.94	6.25	1.01	1.72	5.00	2.85	0.73	1.64
Sweden	1.71	3.92	4.54	6.98	1.36	4.22	0.32	0.78	1.37
Italy	2.66	6.98	7.61	2.35	2.39	3.62	1.32	1.01	0.44
W.Europe	1.19	4.97	6.03	3.33	1.87	4.20	0.98	0.88	1.35
USA	2.34	8.13	8.47	2.96	3.48	4.93	1.67	1.53	2.23
Canada	1.05	7.21	9.16	3.57	3.17	5.46	1.15	1.80	2.27
N.Zealand	2.69	6.93	7.96	4.35	4.30	5.27	0.70	1.49	2.52
Japan	0.68	1.51	2.74	6.65	0.91	0.74	0.05	n/a	n/a
Poland	0.61	5.42	4.90	4.23	2.25	7.97	2.08	1.95	1.89
Slovakia	0.93	3.48	2.39	6.37	3.37	5.69	1.60	2.39*	1.44*
Czech Rep.	0.48	4.82	4.43	3.84	4.78	7.34	0.84	2.39*	1.44*
Overall	1.37	5.26	6.39	3.48	2.20	4.26	0.96	1.05	1.58

Source: van Dijk and Mayhew, Criminal Victimisation in the Industrialised World, 1992
* Combined Czechoslovakia data

The International Crime Victims Survey concluded that, as in all other comparable countries, households in Australian cities of over a million people were over twice as likely as those living in medium sized towns and four times as likely as those living in rural areas to have installed a burglar alarm. They are sixty two per cent more likely than the others to have special locks on their doors, twenty six per cent more likely to have a burglar proof fence around their property and seventeen per cent more likely to have special locks or grilles on their windows. They are also thirty seven per cent more likely to feel very unsafe when walking alone after dark. (See Table 4) Those who live in settlements under 10,000 population are twice as likely as big-city dwellers to feel very safe walking alone after dark.

Big-city dwellers believe, apparently correctly, that urban crime in Australia is worse than crime in less urbanised environments. The risk of burglary in cities of over one million people is almost three times that of towns and rural areas under ten thousand population. The city-dwellers' risk of crimes against the person, including assaults and robberies is also around twice that of the rural population. The fears of the city folk are therefore supported by the statistics on crime frequency. Interestingly, however, the cities between half and one million generally appear to have the worst problems - those which grew fastest during the 1970s and 1980s - rather than the cities over one million population.

Australian cities are not unique or even particularly seriously afflicted in this respect. United States figures on crimes reported to police show that in 1988 the cities of Chicago, Houston, Los Angeles, New York Philadelphia and Washington had rates of reported crime ranging from 2.5 to seven times the national average for Homicide, 2.4 to five times the national average for Robbery, and two to three times the national average for motor vehicle theft. (See Table 5) Equivalent 1988 figures for West Germany and England and Wales show rather less excess in the cities over the national averages, with some cities actually reporting less than average crime. The worst city figures for homicides in these two countries were around twice the national average, while the worst for robbery and motor vehicle theft were around four times the national averages. Australian cities' figures again appear to be more in line with the European cities than with the much higher American ones, with levels of violence far below those of the American cities. Even Australian rates of motor vehicle theft appear modest by comparison with U.S. and U.K. cities' rates.

Table 4. Percentages of Respondents using Crime Prevention Measures
by Type of Measure and City Size, Australia 1992

Crime Prevention Measures and Feelings of Safety			City Size
Crime Prevention Measures and Feelings of Safety	<10,000	10,000 <50,000	50,000 <500,000	500,000 to 1 Million	Over 1 Million

Around the House: - percentages using or owning...

A burglar alarm	4.9	7.4	9.2	13.2	20.3
Special door locks	35.8	46.5	49.7	58.3	77.1
Window, door grilles	17.8	29.0	33.1	38.0	37.0
Dog to deter burglars	47.9	39.0	37.7	30.5	33.0
A high fence	16.5	22.5	20.1	27.0	26.4
Own a gun	39.2	18.6	22.0	12.0	8.5
None of these	18.8	18.6	18.5	16.4	9.4

Walking Alone after Dark: - percentages feeling...

Very safe	45.5	33.1	33.0	21.1	21.2
Fairly safe	38.1	36.0	39.9	37.4	42.4
A bit unsafe	11.9	19.1	17.0	20.1	19.1
Very unsafe	4.6	11.8	10.1	21.4	17.3

Table 5. Reported Crime Rates per 100,000 Population
Homicide, Robbery and Motor Vehicle Theft - Australian and Overseas Cities, 1988

Region (Population)	Homicide	Robbery	Motor Vehicle Theft
Australia (16806730)	2.1	50.5	756.8
Sydney (3566428)	2.0	110.9	1206.3
Melbourne (3039100)	5.1	54.6	1011.1
Brisbane (1154391)	2.1	37.7	479.6
Canberra (277926)	1.8	30.6	516.3
England and Wales (50383000)	1.2	62.4	727.9
London (7242090)	2.0	247.6	941.5
Greater Manchester (2578000)	1.9	75.6	1481.6
West Midlands (2617000)	1.7	97.4	1100.6
Merseyside (148000)	1.5	106.3	1176.3
Thames Valley (1940000)	0.9	22.4	601.0
West Yorkshire (2057000)	1.3	40.1	780.7
West Germany (61241700 )	4.1	47.3	113.6
Berlin (2030128)	5.0	127.2	208.0
Bonn (458616 )	2.6	39.9	108.0
Cologne(928309)	5.6	129.5	179.0
Frankfurt (621379)	9.5	215.3	329.0
Hamburg (1596400)	4.8	172.3	469.0
Munich (1206394)	4.2	76.0	141.0
USA (245807000)	8.4	220.9	582.9
Chicago (2994100)	22.0	967.7	1503.4
Houston (1725421)	25.5	582.4	1163.3
Los Angeles (3402342)	21.6	769.5	1685.0
New York (7346352)	25.8	178.5	1632.6
Philadelphia (1657285)	22.4	537.1	1226.6
Washington DC (620000)	59.5	917.6	1392.4

Source: Mukherjee, S.K., & Dagger, D., The Size of the Crime Problem in Australia, Australian Institute of Criminology, 1990

Some Conclusions and Forecasts about Crime in Australia.

Trends in crime in Australia appear to be related to its demography, its wealth and its high level of urbanisation and the opportunities for crime created by the urban lifestyle. Official statistics of crimes reported to the police, as in other countries during the 1970s and 1980s, appear to show a considerable increase over and above the `real' levels of change, as recorded by crime victims surveys. There are various explanations for this, including deliberate efforts by the community to encourage the reporting of crimes (e.g. domestic violence), and the increasing tendency to report crimes in order to validate claims on insurance policies. The upward changes in reported figures should not therefore be seen as indicative of the size of the increase in crime rates. It is clear, however, that they have been interpreted widely as such, by the media and by political interest groups.

Many changes in the criminal justice system, including an increase in the numbers of police per 100,000 population from 178 in 1973/4 to 244 in 1991/2, have been based on this interpretation. The increase in police numbers, itself, would inevitably exert a further upward pressure on crime report figures. In New South Wales major changes in the sentencing legislation designed to lengthen prison sentences were necessitated, it was argued by the state government, partly by the need to deter crime in that state, even though Victoria, with rates of imprisonment only half those of New South Wales, still had a similar crime level.

To complicate the analysis, the economic recession provides a range of other factors to be considered. It occurred as the baby-boom-echoers were leaving school and looking for jobs, delaying their financial independence and greatly enhancing the frustrations normally felt by adolescents anxious to begin life as adults. It has left many thousands of them with time on their hands, and a proportion of them may have turned to crime which, in better economic circumstances, they might not have become involved in. Coinciding, as it has, with the increased promotion of alcohol to young people, through such devices as aggressive sporting sponsorships, this may be ultimately an explosive mixture.

The recent figures confirm that crime trends can be broadly anticipated, using an insightful combination of demographic projections and data on age-specific rates of offence. The logic suggested by this model can be extended, giving us a warning that we should expect real, although temporary increases in the next few years in crimes associated with persons aged in their twenties - perhaps domestic violence and child abuse, perhaps crimes associated with work and business dealings such as credit card abuse. Similar demographic logic warns us to expect an increase in the fear of crime, which is exhibited by the elderly in spite of the fact that they are relatively unlikely to be victims of crime, because the elderly will be an increasingly large and vocal group in Australia. But we should not allow ourselves again to become hysterical while patterns of crime in Australia continue in their present direction - what is needed instead is meaningful data and rational thinking about society and crime.

Written By John Walker
for other articles by this author, visit http://www.ozemail.com.au/~born1820