Australian Institute of Criminology

Skip to content

Results

In this section, the results of the project are presented in five subsections addressing each of the research questions. First, the number of offender trajectory groups that were identified will be reported. Second, the demographic, offence and criminal justice system event characteristics of the trajectory groups will be discussed. Third, the overall cost of and cost per individual in the offender trajectories will be examined. Next, whether some communities were more likely to generate chronic offenders than others and the extent of residential mobility will be explored. Finally, the 10 percent ranked communities that carried the cost burden of chronic offenders will be identified.

Number of offender trajectory groups

The first research question sought to determine how many distinct offender trajectories could be identified in the criminal careers of individuals in the 1990 cohort. Models with two to seven trajectories were created and the BIC and average group membership probabilities for each of the models were examined (see Table 8). The optimal model included five or six groups, as the seven group model had false convergence. The six group model had a higher value for BIC, while the five group model had a relatively high value for BIC and a slightly higher probability of group membership (>0.75). Examination of the form of the trajectories indicated that the six group model split the chronic offender trajectory into two groups, but did not add to interpretation. Consequently, the model with the smaller number of groups was selected for ease of interpretation (Fergusson, Horwood & Nagan 2000). Estimates of each component were examined to ascertain the form (ie cubic, quadratic, linear and intercept terms) of the five trajectories. Most terms were significant at the 0.5 level and all trajectories had a significant cubic term (see Table 9), so all five groups were assumed to be described best by cubic functions because of the possible impacts of truncation at age 20 and incarceration for this older group.

Table 8 Bayesian Information Criterion and average group membership probability of trajectory models
Number of groups BIC (1) BIC (2) AIC Avg. group membership probability
2 -105950.3 -105935.9 -105890.5 0.96
3 -103267.5 -103247.1 -103182.8 0.91
4 -102299.8 -102273.4 -102190.3 0.92
5 -101663.7 -101631.3 -101529.2 0.79
6 -101049.7 -101010.1 -100885.4 0.78
7 -100810.5 -100764.9 -100621.3 0.73
Table 9 Significance of parameter estimates for final trajectory model
Group Parameter p-value
Group one Intercept 0.8998
  Linear 0.5294
  Quadratic 0.2258
  Cube 0.0488
Group two Intercept 0.0000
  Linear 0.0000
  Quadratic 0.0000
  Cube 0.0000
Group three Intercept 0.2789
  Linear 0.0163
  Quadratic 0.0000
  Cube 0.0000
Group four Intercept 0.0000
  Linear 0.0000
  Quadratic 0.0000
  Cube 0.0000
Group five Intercept 0.0000
  Linear 0.0000
  Quadratic 0.0000
  Cube 0.0000

Figure 2 presents the five offender trajectories identified by the model. Individuals in groups one and two offended less frequently, with individuals in group one averaging 2.1 offences (SD=1.4) and individuals in group two averaging 1.9 offences (SD=1.5). Group one peaked during adolescence when individuals were aged 14 to 16 years, while group two had adult onset where individuals were over 17 years of age. These groups accounted for most of the offender cohort, with 29.3 percent of the cohort in group one and 54.9 percent in group two. Group one was labelled ‘adolescent peaking (low)’ while group two was labelled ‘adult onset (low)’. The third group involved early onset and high levels of offending (M=46.9 offences; SD=46.2 offences), with offending peaking when individuals were 15 years of age. This group included three percent of the cohort and was labelled ‘early onset (chronic)’. Group four had adolescent onset when youth were aged 11 to 14 years with moderate offending. On average, each individual in group four was convicted of 11.2 offences (SD=6.2). This group included 10.5 percent of the cohort and was labelled ‘adolescent onset (moderate)’. The fifth group had adolescent onset of offending when individuals were aged 12 or 13 years, with high levels of offending that peaked when individuals were 20 to 21 years of age. On average, individuals in the fifth group were convicted of 35.0 offences (SD=29.7). Only a small proportion of the cohort was in this group (2.2%), which was labelled ‘adolescent onset (chronic)’.

Figure 2 Offences committed by individuals in the offender trajectories by age in years (n)

Characteristics of offender trajectory groups

The second research question sought to determine What are the demographic, offence and criminal justice system event characteristics associated with trajectory group membership? Table 10 presents the demographic characteristics of the offender trajectory groups. Almost one-quarter (24.5%) of the population offended, although one-fifth (20.6%) were in the two low offending groups. Between 75 percent and 80 percent of each trajectory group were male, with the exception of the adolescent peaking (low) group, which comprised nearly 60 percent males. About one-tenth of the two low offending groups were Indigenous, while one-third of the two adolescent onset groups and nearly half of the early onset (chronic) offender group were Indigenous.

Table 10 Demographic characteristics of offending trajectories
Trajectory group Offenders Male Indigenous % of total
n %a n % n % Populationb
G1 Adolescent peaking—low 4,159 29.3 2,394 57.6 479 11.5 7.2
G2 Adult onset—low 7,778 54.9 5,824 74.9 660 8.5 13.4
G3 Early onset—chronic 428 3.0 336 78.5 211 49.3 0.7
G4 Adolescent onset—moderate 1,488 10.5 1,138 76.5 443 29.8 2.6
G5 Adolescent onset—chronic 318 2.2 257 80.8 102 32.1 0.5
Total 14,171   9,949 70.2 1,895 13.4 24.5

a: Percentages may not total 100 due to rounding

b: Total estimated population of 16 year olds in 2006: 57,954 (ABS 2011a)

The number of offences committed by members of each trajectory group and types of offences committed are presented in Tables 11 and 12. Individuals in the two low offending trajectories accounted for 84.2 percent of offenders and 33 percent of offences. Members of the moderate group were 10.5 percent of offenders and were responsible for 23.4 percent of offences. Members of the two chronic groups were 5.2 percent of offenders and committed 43.7 percent of offences.

Table 11 Offences committed by each trajectory group
Trajectory group Offenders Offences
n %a n %a
G1 Adolescent peaking—low 4,159 29.3 8,923 12.5
G2 Adult onset—low 7,778 54.9 14,626 20.5
G3 Early onset—chronic 428 3.0 20,069 28.1
G4 Adolescent onset —moderate 1,488 10.5 16,680 23.4
G5 Adolescent onset—chronic 318 2.2 11,115 15.6
Total 14,171   71,413  

a: Percentages may not total 100 due to rounding

Visual inspection of the data (see Table 12) indicated that members of the adolescent peaking (low) trajectory were more likely than members of the overall offender cohort to have committed theft and related offences and less likely to have committed unlawful entry offences. Members of the adult onset (low) group were more likely to have committed public order offences, offences against justice procedures and dangerous or negligent acts endangering persons. They were less likely to have committed theft and related offences, unlawful entry offences and property damage offences. Members of the two chronic groups were more likely to have committed unlawful entry offences. Additionally, members of the early onset (chronic) group were more likely to have committed theft and related offences and were less likely to have committed public order offences.

Table 12 Offence types committed by trajectory group membersa
ANZSOC offence type G1 Adolescent peaking—low G2 Adult onset—low G3 Early onset—chronic G4 Adolescent onset—moderate G5 Adolescent onset—chronic Total
n % n % n % n % n % n %
Theft and related offences 3,319 37.2 2,564 17.5 7,351 36.6 4,369 26.2 3,048 27.4 20,651 28.9
Unlawful entry with intent/burglary, break and enter 750 8.4 431 2.9 5,111 25.5 2,008 12.0 2,285 20.6 10,585 14.8
Public order offences 1,212 13.6 4,462 30.5 1,249 6.2 2,466 14.8 1,090 9.8 10,479 14.7
Property damage and environmental pollution 1,055 11.8 987 6.7 2,481 12.4 2,003 12.0 1,543 13.9 8,069 11.3
Offences against justice procedures, government security and government operations (excluding breaches) 375 4.2 1,950 13.3 1,107 5.5 1,559 9.3 772 6.9 5,763 8.1
Illicit drug offences 721 8.1 1,385 9.5 600 3.0 1,462 8.8 702 6.3 4,870 6.8
Acts intended to cause injury 568 6.4 698 4.8 741 3.7 1,136 6.8 424 3.8 3,567 5.0
Dangerous or negligent acts endangering persons 197 2.2 1,218 8.3 179 0.9 332 2.0 125 1.1 2,051 2.9
Deception and related offences 119 1.3 388 2.7 459 2.3 318 1.9 700 6.3 1,984 2.8
Miscellaneous offences 239 2.7 117 0.8 278 1.4 370 2.2 135 1.2 1,139 1.6
Weapons and explosives offences 144 1.6 242 1.7 112 0.6 237 1.4 128 1.2 863 1.2
Sexual assault and related offences 159 1.8 94 0.6 168 0.8 179 1.1 38 0.3 638 0.9
Robbery, extortion and related offences 47 0.5 58 0.4 156 0.8 203 1.2 89 0.8 553 0.8
Abduction and related offences 16 0.2 28 0.2 76 0.4 38 0.2 36 0.3 194 0.3
Homicide and related offences 2 0.0 4 0.0 1 0.0 0 0.0 0 0.0 7 0.0
Total 8,923   14,626   20,069   16,680   11,115   71,413  

a: Percentages may not total 100 due to rounding

Note: Shaded indicates that the offence type was included in the assessment of the wider economic and social costs of offending

Table 13 presents the number of criminal justice system events and days supervision based on trajectory group membership. After taking into account the proportion of the cohort that each offender trajectory group comprised, visual inspection of the data indicated that members of the adolescent peaking (low) trajectory were more likely to have been cautioned and were less likely to have had a court appearance. Members of the adult onset (low) trajectory were less likely to have been cautioned, conferenced or to have had a Children’s Court appearance and were more likely to have had a Magistrates Court appearance. Members of the two chronic offender trajectories and the moderate offender trajectory were more likely to have had all criminal justice system events. Members of these three groups were also found to have been sentenced to a higher number of days detention/incarceration and community-based supervision than would have been expected given the proportion of the offender cohort that each group represented.

Cost of offender trajectory groups

The third research question sought to determine the costs of individuals on different offender trajectories. Table 14 presents these costs. Over four-fifths (84.2%) of the cohort were in the two low offending groups, but these groups accounted for less than one-third (30.4%) of total costs. Approximately one-tenth (10.5%) of the cohort were in the adolescent onset (moderate) group, who accounted for 22.4 percent of the costs. Each individual in the moderate group generated a total cost $58,116, with criminal justice system costs accounting for two-thirds (59.9%) of this cost. While 5.2 percent of the cohort was in the two chronic groups, they accounted for 47.3 percent of the total costs. Each individual offender in the chronic groups cost more than three times as much as someone in the moderate group and over 20 times more than individuals in the two low offending groups. On average, each individual in the adolescent onset (chronic) group generated a total cost of $221,602, while each individual in the early onset (chronic) group cost $262,057.

Table 13 Number of criminal justice system events and days supervision based on trajectory group membership
  Trajectory group
G1 Adolescent peaking—low G2 Adult onset—low G3 Early onset—chronic G4 Adolescent onset—moderate G5 Adolescent onset—chronic  
n % n % n % n % n % na %

Criminal justice system events

Cohort members 4,159 29.3 7,778 54.9 428 3.0 1,488 10.5 318 2.2 14,171 100.0
Caution 4,753 48.5 1,646 16.8 709 7.2 2,256 23.0 435 4.4 9,799 100.0
Police referred conference 323 32.8 68 6.9 136 13.8 372 37.8 85 8.6 984 100.0
Children’s court appearanceb 698 11.3 328 5.3 2,187 35.3 2,120 34.2 866 14.0 6,199 100.0
Magistrates court appearance 855 5.4 9,571 60.0 976 6.1 3,318 20.8 1,239 7.8 15,959 100.0
District court appearance 13 2.8 118 25.1 80 17.0 165 35.0 95 20.2 471 100.0
Supreme court appearance 4 9.3 18 41.9 7 16.3 8 18.6 6 14.0 43 100.0
Total events 6,646 19.9 11,749 35.1 4,095 12.2 8,239 24.6 2,726 8.1 33,455 100.0

Number of days sentenced

Youth detention 2,385 3.0 2,545 3.2 49,673 62.0 14,170 17.7 11,408 14.2 80,181 100.0
Adult incarceration 6,010 3.3 26,483 14.4 41,489 22.5 52,184 28.3 58,357 31.6 184,523 100.0
Youth community-based supervisionc 24,580 4.1 14,720 2.5 283,740 47.7 168,059 28.3 103,151 17.4 594,250 100.0
Adult community-based supervision 18,716 4.1 156,795 34.1 54,950 11.9 157,191 34.2 72,542 15.8 460,194  

a: Percentages may not total 100 due to rounding

b: Children’s court includes Children’s Court and Children’s Court of Queensland

c: Assessed as the most serious outcome for the finalisation

Table 14 Criminal justice system and wider economic and social costs of offender trajectoriesa
Group Cohort members Justice system costs Wider economic and social costs Total costs
n % Mean ($) Group costs ($ mil) % cost Mean ($) Group costs ($ mil) % cost Mean ($) Group costs ($ mil) % cost
G1 Adolescent peaking—low 4,159 29.3 4,127 17.16 8.5 5,408 22.49 12.2 9,535 39.66 10.3
G2 Adult onset—low 7,778 54.9 5,695 44.30 22.0 4,275 33.25 18.0 9,971 77.55 20.1
G3 Early onset—chronic 428 3.0 130,520 55.86 27.7 131,537 56.30 30.4 262,057 112.16 29.0
G4 Adolescent onset—moderate 1,488 10.5 34,780 51.75 25.7 23,337 34.73 18.8 58,116 86.48 22.4
G5 Adolescent onset—chronic 318 2.2 101,497 32.28 16.0 120,106 38.19 20.6 221,602 70.47 18.2
Total 14,171   14,209 201.35   13,052 184.96   27,261 386.31  

a: Percentages may not total 100 due to rounding

The extent that communities generated chronic offenders and the level of residential mobility

The fourth research question sought to determine whether some communities were more likely to generate chronic offenders. For the purposes of this analysis, offenders were classified as chronic if they had been identified in the moderate or chronic offender trajectories (15.7% of the offender cohort). Figure 3 Proportion of chronic offenders by Queensland postal areas Figure 3 graphically displays the proportion of the 16 year old population in each POA identified as chronic offenders. It is evident that chronic offenders are not randomly distributed geographically. About two-thirds of POAs (n=224; 68.1%) had none or a low proportion of the population that were chronic offenders. One-fifth (n=72; 21.9%) of locations were found to have a high proportion of the population who were chronic offenders, where between 5.7 and 9.1 percent of the population were chronic offenders. One-tenth (n=33; 10.0%) had a very high proportion of the population who were chronic offenders, where over nine percent of the population were chronic offenders.

Figure 3 Proportion of chronic offenders by Queensland postal areas

The POAs were then ranked based on the proportion of the population who were chronic offenders. Figure 3 presents the top 10 percent POAs where over nine percent of the population were chronic offenders. While these 33 locations represents 10 percent of all POAs with over 10 individuals aged 16 years old at the time, they accounted for 458 (20.5%) of all chronic offenders. Also presented in this Table is the percentage of 16 year olds in the POA who were Indigenous, the Index of Relative Socio-economic Disadvantage (IRSD) decile and the Australian Standard Geographical Classification—Remoteness Areas. The IRSD is an index developed by the ABS (2006b) that summarises census data about low income, high unemployment and low levels of education. The index scores are presented as deciles, that is, an index score of 1 indicates the postal area is in the 10 percent of most disadvantaged areas in Australia. The Australian Standard Geographical Classification—Remoteness Areas (ABS 2011b) classifies areas into five broad geographical categories based on access to goods and services. These categories include Major Cities, Inner Regional, Outer Regional, Remote and Very Remote.

Examination of the information presented in Table 15 indicates that the majority of these POAs had a high proportion of Indigenous 16 year olds. Twenty-two of the 33 POAs had higher than average (5.38%) populations of Indigenous 16 year olds. In three of these POAs (4713, 4830 and 4876), 100 percent of the 16 year olds were Indigenous. These POAs were also classified by high levels of disadvantage. Ten (30%) were classified as being in the lowest decile of disadvantage (mean=3.36). However, a substantial number of POAs with high proportions of chronic offenders were classified as not disadvantaged. When these POAs are examined, they include the Brisbane City central business district, the inner suburbs of Brisbane and the coastal suburbs around Cairns. A substantial number of POAs (13 of the 33) with high proportions of chronic offenders were classified as remote and very remote. These are areas where it is difficult and costly to deliver goods and services. Additionally, 12 POAs were classified as outer regional. Interestingly, one of the very remote POAs (4730) that had a high proportion of chronic offenders had no officially identified Indigenous 16 year olds (based on the census data) and was not classified as disadvantaged (IRSD decile=6). This POA was in western Queensland and included Longreach.

Table 15 Postal areas with the highest proportion of chronic offenders
Postal area % 16 year old population Indigenous IRSD decile ASGC-RS
4713 100.0 1 Remote Australia
4890 62.5 1 Very Remote Australia
4000 0.0 9 Major Cities of Australia
4824 29.0 3 Remote Australia
4605 45.5 1 Outer Regional Australia
4490 59.1 1 Very Remote Australia
4714 26.1 1 Outer Regional Australia
4830 100.0 2 Very Remote Australia
4465 25.0 2 Remote Australia
4470 10.0 4 Remote Australia
4849 0.0 2 Outer Regional Australia
4387 0.0 2 Outer Regional Australia
4874 55.8 2 Very Remote Australia
4852 0.0 6 Outer Regional Australia
4032 5.0 6 Major Cities of Australia
4876 100.0 1 Very Remote Australia
4825 56.9 4 Remote Australia
4730 0.0 6 Very Remote Australia
4183 27.6 2 Inner Regional Australia
4877 23.5 7 Outer Regional Australia
4888 20.0 1 Outer Regional Australia
4871 53.3 1 Very Remote Australia
4021 3.2 3 Major Cities of Australia
4614 12.0 1 Inner Regional Australia
4880 17.5 2 Outer Regional Australia
4895 46.3 1 Remote Australia
4558 2.4 4 Major Cities of Australia
4355 0.0 3 Inner Regional Australia
4012 0.0 7 Major Cities of Australia
4390 10.1 5 Outer Regional Australia
4814 10.7 7 Outer Regional Australia
4878 7.1 7 Outer Regional Australia
4612 0.0 6 Outer Regional Australia

Note: POA locations are provided in Appendix 1

The fifth research question sought to assess the extent of residential mobility among chronic offenders. On average, each chronic offender had 17.7 (SD=19.5) valid POAs recorded. The number of times that chronic offenders changed postal areas is presented in Table 16. About one-third (31.7%) of chronic offenders only had one POA, while about 32.1 percent had three or more POA changes. Hence, chronic offenders appear to be substantially mobile in terms of the number of times they change residential address after their initial contact with the criminal justice system.

Communities carrying the cost burden of chronic offenders

The sixth research question sought to determine which communities carried the cost burden of chronic offenders. Table 17 presents the top 10 percent of POAs identified based on the total cost to the community of chronic offenders; these are graphically presented in Figure 4. When aggregated and totalled, chronic offenders in each POA were found to cost between $2.4 and $14.0m. Despite representing 10 percent of POAs, the top 33 POAs accounted for 40.4 percent of the chronic offenders, 47.0 percent of offences committed by chronic offenders, 50.5 percent of the total cost of chronic offenders and 35.2 percent of the total cost of the all offenders in the cohort. These areas differed from the areas with the highest proportion of chronic offenders as these estimates do not take into account total population. Consequently, these POAs have the highest number of chronic offenders but not necessarily the highest concentration of chronic offenders.

A different picture emerged when the costs of chronic offending were examined (see Figure 4). Regional Queensland appears to be carrying the major cost burden of chronic offenders. Almost half of the high-cost POAs were classified as regional. The POA that incurred the highest cost of chronic offenders was 4350, with the cost estimated at over $14m dollars. This POA includes the regional city of Toowoomba. Only three of the areas were classified as remote or very remote. These POAs had high proportions of Indigenous young people and high levels of disadvantage. The cost of crime in these areas is considerable.

Figure 4 Distribution of total costs for chronic offenders by Queensland postal areas

Finally, both the concentration of chronic offenders (Top 10% proportion of population chronic offenders) and the cost of chronic offenders (Top 10% total cost of chronic offenders) were mapped to examine the spatial distribution of these postal areas (see Figure 4). Eight POAs were identified that experienced high concentrations of chronic offenders and high costs of chronic offenders. These POAs are predominantly located in north and far north Queensland and contain a high proportion of Indigenous young people. The outer regional POA in Inner South West (Insert D of Figure 4) includes Cherbourg—a large Indigenous community. This map also clearly indicates that the costly POAs include the outer suburbs of Brisbane and the regional areas of Rockhampton, Gladstone and Toowoomba. However, the areas where high rates of chronic offenders are located are predominately in the remote and very remote areas of Queensland.

Figure 5 Top 10 percent of locations based on proportion in population who were chronic offenders and the top 10 percent of most costly high-rate offender postcodes