Sentencing Indigenous and Non-Indigenous Women in Western Australia's Higher Courts

This article presents results from an exploratory study seeking to examine the role of sentencing in the continuing overrepresentation of Indigenous women in Western Australia's prisons. Sentencing data from Western Australia's higher courts indicate that Indigenous women were less likely than non-Indigenous women to be sentenced to a term of imprisonment when appearing before the court for comparable offending behaviour and histories.

Indigenous female offenders are grossly overrepresented in Western Australia's prisons. The number, proportion and rate of Indigenous women imprisoned in this jurisdiction have increased substantially since the late 1990s. The Royal Commission into Aboriginal Deaths in Custody (1991) made numerous recommendations to reduce the number of Indigenous people imprisoned in Australia, stressing that imprisonment for Indigenous offenders should be the option of last resort. Steady increases in the number, proportion and rate of Indigenous women imprisoned in Western Australia, and the associated overrepresentation of Indigenous women in prisons, is therefore a significant issue. This article presents the results of an exploratory study comparing the sentencing outcomes of Indigenous and non-Indigenous women in Western Australia's Higher Courts, in order to understand what role (if any) disparate treatment at sentencing might play in this concerning trend.

Indigenous Women and Imprisonment in Western Australia
Between 1998 and 2007, the number of Indigenous women prisoners in Western Australia increased by 142%, from 72 in 1998 to 174 in 2007. In 1998, Indigenous women comprised 47% of the Western Australian female prison population. By 2007, this proportion had increased to 54%, even though Indigenous peoples constitute only 3.8% of Western Australia's estimated resident population (Australian Bureau of Statistics 1998Statistics , 2006bStatistics , 2007.
In relative terms, the rate of Indigenous women's imprisonment in Western Australia increased from 439.7 per 100,000 population in 1998, to 789.2 in 2007. This compared with an overall increase in female imprisonment rates from 22.7 in lian Bureau of Statistics, 1998Statistics, -2007. In this jurisdiction, Indigenous female imprisonment rates have consistently exceeded national rates. Aside from an anomalous figure in 2002, Western Australian rates have remained around double those recorded nationally (Figure 1) (Australian Bureau of Statistics, 1998-2007.
A key question is whether the overrepresentation of Indigenous women in this jurisdiction is due to differences in criminal offending and past history of contact with the criminal justice system, or to differences in treatment at the point of sentencing.

Indigenous Women and Sentencing in Western Australia
Judicial practices at sentencing could provide some explanation for the overrepresentation of Indigenous women in Western Australia's prisons. The Royal Commission into Aboriginal Deaths in Custody (1991) argued that ''the powers and decisions of sentencing courts present considerable opportunity for reducing the numbers of Aboriginal people in custody''. Conversely, sentencing practices may aggravate sentencing outcomes for Indigenous offenders.
As is the case nationally, data by Indigenous status and sentencing outcome are published only sporadically in Western Australia. While gender breakdowns are not provided, these data show that Indigenous offenders are more likely to be sentenced to imprisonment. For example, in the most recently available figures, 66.3% of Indigenous offenders sentenced in Western Australia's Higher Courts during 2001 were given a term of imprisonment compared with 60.3% of non-Indigenous offenders (Loh & Ferrante, 2003).
This finding is suggestive of discriminatory sentencing practices favouring non-Indigenous over Indigenous offenders. These sentencing outcomes, however, may simply be a response to differences in offending behaviours (Weatherburn, Fitzgerald, & Hua, 2003). Sentencing research indicates that there is a strong correlation between the seriousness of the offender's criminal history, the severity of the offender's crime(s) and sentencing outcomes. Offenders exhibiting more extensive and more serious forms of criminality tend to receive harsher sentences (Mitchell, 2005;Spohn, 2000). Further, given that around 85% of all sentenced offenders in Western Australia are male, any discrepancies highlighted in these data arguably denote sentencing differences between Indigenous and non-Indigenous men and offer little insight with regard to women (Government of Western Australia, 2005).

Current Study
Using Higher Court (District and Supreme) data supplied by the Western Australian Crime Research Centre, the current study examines whether Indigenous women were more likely than non-Indigenous women to receive a sentence of imprisonment for comparable offending behaviour and histories.

Methods
To investigate the possible role of sentencing disparity in the rising Indigenous female prison population, imprisonment outcomes were investigated over a 9-year period from 1996 to 2005. During this time a total of 3,133 cases resulted in the sentencing of a female defendant, and 28.6% of these involved an Indigenous woman (N ¼ 896). Due to missing information, however, in some cases, 2,789 cases were used in our final analyses.
Logistic regression analyses were used to estimate the separate independent impact of Indigenous status on the imprisonment sentencing decision (dependent variable), while controlling for the effect of age at sentencing, seriousness of principal offence, presence of a guilty plea, presence of multiple conviction counts, and extensiveness of prior arrests between 1984 and 2005. A description of the variables and their coding can be found in Table 1.
Sentencing decisions are affected by judicial perceptions of an offender's ability to ''do time'', but can also be influenced by criminal stereotyping (Johnson, 2003;Steffensmeier, Ulmer, & Kramer, 1998). Age can be important in deciding whether a sentence of imprisonment may be unduly harsh. For example, younger people could be perceived as less able to cope with imprisonment (Wu, 2007). At the same time, and somewhat paradoxically, criminality (or the threat of criminality) is often associated with youthfulness (Steffensmeier et al., 1998;White & Perrone, 2005). This means that being younger could potentially increase the chances of imprisonment. Because the Indigenous population has a younger age structure than the non-Indigenous population, we need to consider the effect of age on the decision to imprison.
When making sentencing decisions, judges must impose punishment proportionate to the criminal harm caused, while also taking into consideration how culpable offenders are for their actions and any potential risks they pose to the community (Steffensmeier et al., 1998;White & Perrone, 2005). Current crime seriousness, criminal history, and pleading guilty are crucial to these considerations. In sentencing research, these factors are generally the strongest predictors of sentence severity (Mitchell, 2005;Spohn, 2000). In Australia, offence seriousness is codified in law using statutory offence classifications and prescribed sentencing penalties. In 1997 the Australian Bureau of Statistics developed the Australian Standard Offence Classification System, a numerical codification and systematic ordering of offences defined, either implicitly in the common law or explicitly in Australian legal codes (Australian Bureau of Statistics, 1997). The National Offence Index was developed by the Australian Bureau of Statistics, and ranks all offence classifications contained within the Australian Standard Offence Classification System in order of seriousness, from 1 to 155, with 1 being the most serious and 155 being the least serious. To assist in the interpretation of the findings, this was reverse-coded. The number of conviction counts was included as an additional measure of current crime seriousness, while the measure of prior arrests indicates extensiveness of an offender's criminal history.
Finally, if Indigenous status has an impact on the decision to imprison, after adjusting for differences in offending and prior criminal histories, then research suggests that this may be due to criminal stereotypes or attributions. Theories of sentencing disparities argue that differing perceptions of motivations for offending, remorsefulness, and responsibility may be attributed to defendants of differing ethnic groups (Jeffries & Bond, 2009).

Results
We first examined any differences between Indigenous and non-Indigenous offenders in the decision to imprison and each independent variable over the 9-year study period (1996)(1997)(1998)(1999)(2000)(2001)(2002)(2003)(2004)(2005). Sentenced Indigenous women were on average slightly younger than non-Indigenous women. The mean age for Indigenous women was 27.6 years compared with 29.4 years for non-Indigenous women. Overall, Indigenous women's past offending was more serious than their non-Indigenous counterparts. On average, Indigenous women had more prior arrests (M ¼ 22.3) than non-Indigenous women (M ¼ 7.7). The mean seriousness of offence score for Indigenous women was 61.5, while that for the non-Indigenous women was 59.5, with higher scores indicating less serious offences. Approximately one-quarter of Indigenous women (24.7%) had multiple conviction counts, compared with 34.1% of non-Indigenous women. A larger proportion of Indigenous women, however, pleaded guilty (80.6%) than non-Indigenous women (67.6%). Finally, a slightly higher proportion of Indigenous women received a prison sentence (49.7%) than did non-Indigenous women (46.4%).
We conducted a series of logistic regressions (by year) to establish whether Indigenous status directly influenced the likelihood of imprisonment when women of like age and offending histories were sentenced for comparable crimes after entering the same plea. Adjusting for the influence of age, seriousness of principal offence, presence of conviction counts, presence of a guilty plea and prior arrests, differences between Indigenous and non-Indigenous women's sentencing outcomes were found. Figure 2 reports the odds ratios for each of the variables in our analyses. (More detailed results are provided in Appendix A.) Depending on the year, the reported models explain between 4% and 19% of the variance in decision to imprison (as estimated by the pseudo R 2 ), suggesting that there may be other variables that may explain this sentencing outcome (such as being on remand at the time of sentence, mental health issues, evidence of premeditation, or indications of remorse). The odds ratios represent the separate impact of each variable, adjusted for the influence of the other factors of interest. Odds ratios allow us to compare the likelihood of a prison sentence for two groups (e.g., Indigenous and non-Indigenous female offenders). An odds ratio of 1 indicates that the likelihood of a decision to imprison is equally likely for both groups. An odds ratio 41 implies that imprisonment is more likely in the first group, while an odds ratio 51 indicates that it is less likely.
As shown in Figure 2, Indigenous women were between 0.41 and 1.20 times less likely than non-Indigenous women to receive a sentence of imprisonment. In others words, Indigenous status was found to have an independent impact on the decision to imprison, but in a somewhat unexpected direction. In all but one year, Indigenous women (compared to non-Indigenous women of like age who plead guilty with similar past and present criminality) were less likely to be sentenced to prison. Overall, on average for the whole period, Indigenous women were 0.70 times less likely to receive a prison sentence. Since 2000, however, the likelihood of imprisonment for Indigenous women, compared to non-Indigenous women under these similar circumstances, has been decreasing. (A pooled model was estimated with dummy variables for each year [with 1996 as the reference category]. The results confirmed that there was a statistically significant effect for time.) The difference in the odds of imprisonment between Indigenous and non-Indigenous women was not statistically significant in all years (Appendix A). That is, the size of the effect could have occurred by chance, rather than be an estimate of a true effect. Caution must therefore be exercised in making strong conclusions at this point. Nonetheless the failure to achieve statistical significance is likely due to the small number of Indigenous female offenders in some years (Appendix A). This pattern of results, however, at least suggests that the effect of Indigenous status may vary over the study period.
The independent impact of criminal history and current crime seriousness on Figure 2. Independent effect of Indigenous status, age, current and past offending behaviour on the decision to imprison by year (Western Australia, 1996. Odds ratios reported. The odds ratio represents the separate effect of the variable of interest, after adjusting for the other factors in the analysis. the likelihood of receiving a prison sentence conformed to expectations. More serious current and past offending histories increased the likelihood of being sentenced to prison. (Although the odds ratio for seriousness of the principal offence is close to 1.0, this is somewhat misleading about the magnitude of the effect. Recall that the measure of offence seriousness ranges from 1 to 155. The odds ratio represents the increase in the odds of a decision to imprison for each unit increase in offence seriousness.) Overall, pleading guilty reduced the odds of a decision to imprison. Older offenders were more likely to receive a sentence of imprisonment, suggesting that leniency is extended to youth regardless of other factors.

Discussion
The current research compared the sentencing outcomes of Indigenous and non-Indigenous women in Western Australia's Higher Courts. Although in some years past and current offending may explain differences in outcomes between Indigenous and non-Indigenous women, there is also evidence that Indigenous women may be less likely than their non-Indigenous counterparts to receive a sentence of imprisonment for comparable criminality. Results suggests (at least for part of the study period) that for Indigenous women in Western Australia's Higher Courts, imprisonment may be considered as an option of last resort. Possible explanations for this finding may rest on the differing social circumstances of Indigenous and non-Indigenous women.
First, as noted earlier, research shows that sentencing decisions are often affected by judicial perceptions regarding an offender's ability to ''do time'' (Mitchell, 2005;Spohn, 2000;Steffensmeier et al., 1998;White & Perrone, 2005). For example, prison may be especially difficult for offenders with social histories marred by violent victimization, mental and/or physical health problems (Birmingham, 2003;Freudenberg, 2001). Indigenous women are more likely than non-Indigenous women to have negative life experiences such as these, thus increasing the possibility of judicial leniency.
Second, in contrast to non-Indigenous women, there may be heightened judicial concern surrounding the social cost (Daly, 1989;Jeffries, 2002a,b) of removing Indigenous mothers (as primary caregivers) from their families. Indigenous women often care for both their own and others' biological children. The incarceration of primary caregivers is therefore especially concerning for Indigenous families and communities, which are often already stretched to the limit of survival (Jones, Masters, Griffiths, & Moulday, 2002).
The above possibilities are likely bolstered by political and community expectations of the criminal justice system after the Royal Commission, and the potential role of sentencing in reducing Indigenous overrepresentation. In line with the Royal Commission, it is possible that the Western Australian judiciary are extending leniency to Indigenous women because they are sensitive to their unique social, economic, political, and historical circumstances. By taking these into account, Indigenous women may be perceived as less blameworthy than their non-Indigenous counterparts, possibly due to Australia's legacy of colonization, associated Indigenous social and economic marginalization, and the potential of imprisonment to further exacerbate this (Jeffries & Bond, 2009).
In recent years a number of Australian jurisdictions have developed alternative ways of sentencing Indigenous offenders (e.g., Indigenous and circle sentencing courts) in an attempt to acknowledge and address the differential needs of Indigenous defendants (Harris, 2004(Harris, , 2006. In case law, recent precedent exists for factors associated with offenders' Indigenous status (e.g., associated disadvantage) and Indigeneity itself (e.g., historical legacy of colonization) to mitigate sentencing (see discussion by Edney, 2003;Edney & Bagaric, 2007) Clearly, the judiciary (as illustrated through Indigenous courts and sentencing precedents) is already attuned to the power that sentencing can play in meeting political and community expectations with regard to reducing Indigenous overrepresentation (Jeffries & Bond, 2009).
Thus, within the above context, it is not unreasonable to suggest that in Western Australia a degree of judicial cognisance may exist around the special circumstances of Indigenous women, and that this in turn may explain why Indigenous women may be less likely than non-Indigenous women to be imprisoned.
Of some concern, however, are the present results suggesting that any special consideration extended to Indigenous women may be dissipating over time. Social, economic, political, and historical differences between Indigenous and non-Indigenous women subsist to the benefit of the latter. Under these unequal circumstances, equitable (rather than equal) treatment is arguably a more just response (Jeffries & Bond, 2009). Equitability, however, may soon become equality, which may add further to the problem of Indigenous overrepresentation.

Directions for Future Research
As of yet we do not fully understand what it is about Indigenous status that is making a difference to judges. Thus, it is crucial that we gain a more thorough understanding of how Indigeneity influences sentencing. In the future, researchers exploring the relationship between Indigenous status and sentencing for women should include social history measures (i.e., victimization, health, childcare responsibilities). This would help to determine what (if any) role these factors play in Indigenous versus non-Indigenous female sentencing. Ideally, prospective studies could also incorporate a qualitative component to their design, including perhaps an examination of judges' sentencing remarks and interviews with judges to better understand judicial sentencing philosophies and practices, and their implications for Indigenous offenders.
Logistic Results (Odds Ratios) of Decision to Imprison on Age, Current and Past Offending Behaviour by Year (Western Australia, 1996 Appendix A