Public opinion on sentencing and parole in Australia

Introduction

There are nine different legislative regimes for sentencing and parole in Australia: one federal system and eight states and territories. In descending order of population, these are New South Wales (NSW), Victoria, Queensland, Western Australia, South Australia, Tasmania, the Australian Capital Territory (ACT) and the Northern Territory. There are eight prison systems, with federal offenders ¹ – who account for about 2 percent of offenders (Australian Bureau of Statistics (ABS), 2017b) – serving their time in state and territory prisons.

As Tubex et al. have discussed, there is wide variation in imprisonment rates and Australia is therefore an ‘ideal penal laboratory’ (2015: 346). Nationally, the rate is 217 per 100,000 head of population, but this ranges from 146 in Tasmania to 880 in the Northern Territory (ABS, 2017b). NSW is the largest jurisdiction in Australia; with around a third of the general and prison population, it tends to drive national imprisonment rates (Tubex et al., 2015).

As at June 2017, there were just over 41,000 prisoners, the highest figure on record, and numbers were rising at approximately 7 percent a year (ABS, 2017b; for comment, see Bartels, 2017; Freiberg, 2016). Indigenous peoples are significantly overrepresented across the Australian criminal justice system, especially in prison (see e.g. Australian Law Reform Commission, 2017; Tubex et al., 2015 for discussion). Although Indigenous peoples account for less than 3 percent of the general Australian population (ABS, 2017a), they comprise 28 percent of the adult prison population (ABS, 2017b). With rising prisoner numbers, the number of persons on parole, or considered to be eligible for parole, is only likely to increase.

While the evidence suggests that the reasons for the rising imprisonment rates in Australia are multifaceted (Tubex et al., 2015; Weatherburn and Ramsey, 2016), recent crises in a number of parole systems across the country have underpinned rises in several states. These crises have arisen as a result of a handful of high-profile parole violations and the recommendations of subsequent parole system reviews and inquiries. Changes to parole systems are often justified on the grounds of a public desire for greater safety, despite a lack of empirical evidence about how the public sees parole. This phenomenon is not unique to Australia; indeed, as Roberts et al. point out, criminal justice policymaking can often occur on the basis of a ‘caricatured understanding of the public’s real views’ (2003: 5).

This paper summarizes the findings from a national study on public opinion on parole in Australia, which represented the first public opinion research in respect of parole, as well as other recent research on public opinion on parole. These findings are situated in the context of recent Australian research on public opinion and sentencing. We also provide an overview of the Australian legislative and policy landscape in respect of parole. The article concludes with some comments on future directions in research and policy.

Australian public opinion on sentencing

Warner et al. (2009) suggested that public attitudes to sentencing matter because of the contribution these attitudes make to public confidence in the criminal justice system; because it is generally accepted that sentencing policy and practice should be responsive to public opinion; and because perceptions of public opinion can force changes to the law. It is clearly important for policy decisions to be based on accurate perceptions. As White and Perrone have noted, however:

…public concerns over sentencing have focused, in particular, on a number of assumptions that generally have little or no basis in empirical reality, and/or fail to take into account the need to balance a number of sentencing objectives. These include perceptions that:

The judiciary is out of touch with public concern and has a propensity to being ‘soft on crime’.

There has been a proliferation of Australian research over the last decade on actual – as opposed to presumed – public opinion on sentencing. In the Melbourne Criminology Study, Victorian judges presented four cases, involving six offenders, to more than 470 members of the public in 32 groups. In the first session with each group, Lovegrove (2007, 2011, 2013) gave a 70-minute talk on sentencing, followed by the judge making general observations about sentencing and taking questions. In a subsequent session, the judge provided an edited version of his sentencing judgment, as well as information about the legislative maximum for the offence(s) and sentencing practices. The participants then wrote down what they considered to be the appropriate sentence for the offence, before being provided with the actual sentence imposed by the judge. Lovegrove (2007) found that although respondents described judicial sentencing as ‘too soft’, the median sentences they imposed were more lenient than those imposed by the judge for three of the four male offenders, while the judge’s sentence was marginally lower in the remaining case. In the two cases involving female offenders, the participants imposed the same sentence type, but a lower median fine level in one case. In subsequent research, Lovegrove explored participants’ views on relevant mitigating factors (2011) and the theoretical explanations for these views (2013).

In another major project (the National Sentencing Study), a telephone survey with 6005 people across Australia found that most respondents expressed high levels of punitiveness and were dissatisfied with sentences imposed by the courts. This is consistent with other ‘top-of-the-head’ findings (see Gelb, 2006, for discussion). Notably, Mackenzie et al. (2012) found that 59 percent of respondents felt that the sentences handed down by courts were too lenient. Respondents were more likely to say sentences were too lenient for violent crimes (79 percent) than for non-violent property crimes (61 percent) or non-violent drug offences (51 percent). However, many were also supportive of the use of alternatives to imprisonment for a range of offences. For example, 82 percent supported the use of alternatives to prison for mentally ill offenders, while 80 and 66 percent respectively supported the use of alternatives for young and drug-addicted offenders. In addition, 55 percent agreed that fewer prison sentences should be given to non-violent offenders (Mackenzie et al., 2012). ²

Analysis by jurisdiction found that there were only minor differences in the key measures of public attitude despite the notable interjurisdictional differences in relation to sentencing policy (Roberts et al., 2011). Differences in public attitudes across jurisdictions accounted for less than 2 percent of variation in confidence in sentencing and punitive attitudes scores. In addition, lack of confidence in sentencing and punitiveness were not related in any systematic way to jurisdictional differences in imprisonment rates. It follows from these findings that the wide differences in sentencing practice and policy between jurisdictions in Australia are not linked to differences in public attitudes. This suggests that sentencing policy is better understood as a function of political initiative, rather than a direct articulation of public attitude.

The Victorian Sentencing Advisory Council (VSAC) undertook additional research on the National Sentencing Study with 1200 Victorian residents. This revealed that there was clear support for using alternatives to prison as a way of addressing the increasing number of people in prison and the subsequent prison overcrowding (Gelb, 2011a). Specifically, 51 percent of respondents agreed that ‘we need to find alternatives to prison to reduce the high cost to the community of keeping people in prison’ was ‘very important’, while 69 percent said it was ‘very important’ that ‘taxpayer money should be used on programs that reduce crime in the first place rather than on prison’. Perhaps most surprisingly, 76 percent said that the argument ‘prisons should be used mainly for dangerous and violent offenders’ was very important in relation to prison overcrowding. When respondents were faced with a forced choice between two approaches, 74 percent favoured ‘increasing the use of alternatives to imprisonment’ as their final choice, while 26 percent chose to ‘build more prisons’. Alternatives to prison were most supported for offenders who were mentally ill (92 percent), young (88 percent), drug-addicted (84 percent) and non-violent (75 percent). In addition, 84 percent agreed that drug-addicted offenders should be put on intensive rehabilitation and counselling programs, rather than being sent to prison, while 75 percent agreed that non-violent offenders should receive community corrections orders rather than prison.

Other findings from the VSAC’s additional research on this project indicated that although respondents said that they were not satisfied with the decisions that courts make (57 percent disagreed, while 28 percent agreed), they nevertheless had some confidence that courts impose an appropriate sentence most of the time (54 percent agreed, vs 40 percent disagreeing) (Gelb, 2011b). Respondents were also asked to nominate the most important purpose of sentencing for a range of offence and offender types (Gelb, 2011c). For first-time burglars, respondents were most likely to nominate rehabilitation as the most important sentencing purpose for both young (62 percent) and adult (50 percent) offenders, followed by punishment (13 percent and 23 percent respectively). For repeat burglars, by contrast, the most important purpose was punishment for both young and adult offenders (40 percent and 51 percent), followed by rehabilitation for young offenders (26 percent) and incapacitation for adult offenders (20 percent). For first-time serious assault offenders, the most important purpose was rehabilitation, nominated by 56 percent and 38 percent of respondents for young and adult offenders respectively. This was followed by punishment (19 percent and 34 percent respectively). For repeat offenders, punishment was the most popular purpose for both young (44 percent) and adult (51 percent) offenders, followed by incapacitation (23 percent and 30 percent respectively).

These findings are consistent with follow-up research on the National Sentencing Study by Spiranovic et al. (2012), with 800 randomly selected respondents across Australia. This indicated that rehabilitation was seen as most important for first-time, young and burglary offenders, while punishment was endorsed as most important for repeat, adult and serious assault offenders. In addition, the odds of choosing rehabilitation compared with punishment were significantly increased by a factor of 6.1 for cases involving first-time offenders.

Other research has indicated that members of the public in NSW were equally disposed to reducing crime by rehabilitating offenders as they were to imprisoning them for longer (Jones and Weatherburn, 2011). Simpson and Butler (2015) recently found that citizens’ juries in NSW, Western Australia and the ACT (n=43; Simpson et al., 2015) demonstrated a preference for retaining deprivation of liberty for very serious offences, but there was also strong support for non-punitive approaches and alternatives to incarceration, as well as better services and programmes to address the underlying determinants of crime (e.g. mental and physical health services and substance abuse treatment: Simpson et al., 2015); prison diversion programmes; raising awareness of prison alternatives; and a commitment to allocating public funds to non-incarceration options. Simpson et al. (2015) also reported that the juries identified equity and fairness, a prevention focus, and community involvement as principles that should underpin offender treatment.

Another significant project was the Tasmanian Jury Sentencing Project, led by Warner. This was the first study of its kind in the world, as it involved surveying and interviewing jurors – as informed members of the public – to determine their views on sentencing generally and the specific outcome in the case in which they had served as jurors (see Warner et al., 2009). The study involved 257 jurors from 51 trials at the first stage (where respondents were asked to complete a consent form and short questionnaire); 144 at the second stage (after respondents had been provided with a booklet about crime and sentencing patterns and the judicial officer’s comments on sentence in the case in which they had served as jurors) and follow-up interviews with 63 respondents. The preliminary findings indicated that 50 percent of respondents imposed a sentence that was less severe than that imposed by the judge, 46 percent were more severe and about 5 percent suggested a sentence of the same severity. In cases involving sex offences, the juror’s proposed sentence was more severe than the judge’s in 53 percent of cases; in property cases, by contrast, the juror’s proposed sentence was less severe than that proposed by the jurors in 66 percent of cases.

The final results from the Tasmanian Jury Sentencing Project involved 698 jurors from 138 trials out of a possible 162 trials that returned a guilty verdict in the two-year study period (Warner et al., 2011). This indicated that more than half of the jurors surveyed suggested a more lenient sentence than the trial judge imposed. Moreover, when they were informed of the sentence imposed, 90 percent said that the judge’s sentence was appropriate. As Warner et al. noted:

The fact that 52 percent of jurors chose a more lenient sentence than the judge and only 44 percent were more severe than the judge shows that informed members of the public are not as punitive as many representative surveys have suggested. (2011: 5) ³

Sentences for sexual offences in Victoria were also the focus of a 2015 study by Devilly and Le Grand. This involved a representative sample of 115 people who were asked to rate four sexual assault vignettes in relation to sentencing outcomes. This study demonstrated that the public’s sentences were very similar to the actual sentence imposed by the judge in three cases, while the fourth was more lenient (a mean imprisonment length of 16.5 years, compared with the judge’s sentence of 25 years). The authors concluded that their findings were ‘inconsistent with the notion that “judges are soft on crime”‘ (2015: 195).

A national Australian study on jurors’ attitudes to sentences for sex offences is currently underway (see Bartels et al., 2014; Warner, 2014). ⁵ The key research questions for this study include whether informed members of the public consider sentencing for sexual offences to be too lenient; whether there are differences in perceptions on the basis of the type of sexual offence; and whether there are any jurisdictional differences in perceptions of leniency for such offences. Analysis of the data in this study is expected to commence in 2018.

While sentencing research is more established in Australia, gathering empirical information on public attitudes toward parole has become germane as a result of recent events with respect to parole. We now turn to this context.

The Australian legislative and policy parole landscape

Parole is a form of conditional release of offenders sentenced to a term of imprisonment which allows an offender to serve the whole or part of their sentence in the community, subject to conditions. It plays a significant part in the Australian criminal justice system. In June 2017, there were 14,871 persons on parole across Australia (ABS, 2017b). With rising prisoner numbers, the number of persons on parole, or considered to be eligible for parole, is only likely to increase.

Almost all of Australia’s eight states and territories have some form of legislatively established independent parole release authority (most commonly known as parole boards), usually comprised of retired judicial officers, full-time members of the parole authority and community members reflecting a wide range of interests and skills. There is no parole board in relation to federal offences and the final decision as to release rests with a political officer, the Federal Attorney-General.

Depending on the jurisdiction, parole release may be automatic at the end of a legislatively specified period (generally only for shorter sentences: see Bartels, 2013; Freiberg, 2016), or at the discretion of a parole board, at the expiration of a minimum term or non-parole period set by a court. The High Court of Australia has described the latter as the ‘minimum period of imprisonment to be served because the sentencing judge considers that the crime committed calls for such detention’ (Power v The Queen, 1974: 628). Automatic court-ordered parole is currently available in three jurisdictions (NSW, Queensland and South Australia). In an inquiry on Indigenous incarceration rates, the Australian Law Reform Commission recently proposed that all states and territories adopt this model (2017: Recommendation 9-2).

Court-imposed non-parole periods tend to range from 50 to 80 percent of the total sentence imposed, but in an increasing number of instances, legislatures have become more prescriptive in their approach to the imposition of such terms, setting them at half or more of the total sentence imposed by the court. Legislatures have also introduced mandatory or presumptive non-parole periods. Under some schemes, the length of the non-parole period is specifically prescribed in legislation (e.g. a mandatory minimum non-parole period of 20 years for murder in South Australia), while in others a set percentage of the head (or total) sentence must be served before the offender is eligible for parole (e.g. at least 80 percent of the total sentence for serious violent offences in Queensland).

A number of high-profile cases of offending on parole have precipitated extensive changes in some parole systems. Most notably, in September 2012, a young journalist, Jill Meagher, was raped and killed while walking home from a night out in Melbourne, Victoria, by Adrian Bayley, who was on parole for a series of sex offences, many against sex workers (Bartels, 2013). This case received extensive national and international media attention and gave rise to a major review of Victoria’s parole laws by a former High Court judge, Ian Callinan QC. The ensuing report was highly critical of the Victorian parole system and prompted a series of punitive legislative and policy responses (Callinan, 2013; for discussion, see Bartels, 2013; Freiberg, 2016). In fact, the relevant legislation, the Corrections Act 1986 (Vic), was only amended once in relation to parole between 2004 and 2012, compared with 13 amendments between 2012 and 2017. Although Victoria had traditionally had one of the lowest imprisonment rates in Australia, it has been estimated that the changes to parole eligibility added some 600 people to its prison population (see Tubex et al., 2015).

In 2016, an elderly woman in Queensland was allegedly killed by a man who had been released on parole only hours earlier, prompting the Queensland Government to commission a review of its parole system by former Solicitor-General Walter Sofronoff QC. The Sofronoff report (2016) made 91 recommendations, all but two of which the Queensland Government (2017) accepted. This report was much more temperate than the Callinan review and generally affirmed the value of parole.

In July 2017, Victoria experienced a further parole crisis when a man took a member of the public hostage and killed another, as well as shooting three police officers. He was subsequently shot dead by police. When it emerged that he was on parole and had previously been charged with but acquitted of terrorism-related offences in NSW (i.e. a separate jurisdiction), the Prime Minister questioned why the man had been released on parole and called for national parole laws (Hunter, 2017; for comment, see Sarre and Bartels, 2017). It should be noted that this is very unusual; as parole is almost entirely a matter for the states and territories, the Prime Minister would not usually get involved in such issues. A few days after this case, the leaders of all nine Australian governments met and agreed to introduce a presumption against parole for persons who have demonstrated support for, or had links to, terrorist activity (Council of Australian Governments, 2017).

Taking into account the foregoing developments, Freiberg (2017) and Freiberg et al. (2017) have identified five key themes in the recent reforms to Australian parole laws and policies. Firstly, there is prioritisation of community safety over all other relevant considerations in parole decision-making, especially in relation to specific offender groups presumed to be most dangerous to public safety, such as sexual and/or violent offenders. Secondly, more limits are being imposed on the courts’ discretion to set non-parole periods, through the use of mandatory or presumptive non-parole periods. Inferentially, courts are becoming less trusted by legislatures to choose an ‘appropriate’ punishment for the crime. The third theme is the questioning or undermining of the discretion of parole boards to make decisions involving perpetrators of serious offences. In such cases, parole board decisions may be reviewed by additional decision-making authorities or are required to be made by differently constituted parole boards.

The fourth theme is the elevation of victims’ rights. This is most evident in the rapid adoption since 2015 of ‘no body, no parole’ laws, which prevent offenders convicted of murder or manslaughter from being released on parole if they have not provided information on the whereabouts of their victims’ bodies. These laws are now in place in Victoria, Queensland, South Australia and the Northern Territory. NSW and Western Australia have also indicated their support for such laws. Finally, there appears to be a shift in the rationale for parole from a prisoner-centred and reintegrative process to one that is increasingly focused on a prisoner’s progressive forfeiture of rights due to their offending behaviour. Convicted criminals are therefore regarded as having intentionally repudiated their contract with society (Fitzgerald et al., under review).

According to Freiberg and others (Freiberg, 2017; Freiberg et al., 2017), Australian governments have too often succumbed to perceived community pressure to restrict parole and the independence and powers of parole authorities. Arguably, parole, like sentencing, should remain in the hands of impartial and independent bodies free of political influence. It therefore behoves legislatures to recognise that an effectively functioning parole system provides the community with a valuable mechanism for promoting community safety.

Australian public opinion on parole

This section summarises the key findings from the first results of a national study on public opinion on parole (see Fitzgerald et al., 2016) and a more recent Western Australian study (Gately et al., 2017).

Fitzgerald et al. (2016) conducted telephone interviews with a nationally representative sample of 1200 Australians in all states and territories. They also undertook in-depth follow-up interviews with 30 respondents, who were presented with a series of parole vignettes. The data for this study were gathered after the Bayley case described above, but before the more recent cases in Queensland and Victoria.

Fitzgerald et al. found mixed support for parole, with 46 percent of respondents agreeing that ‘prisoners should be released to serve the last part of their sentence in the community under supervision’, while 38 percent disagreed. However, there was also support for ‘truth in sentencing’, with 59 percent agreeing that prisoners should serve their ‘entire sentence’ in custody; by contrast, only 27 percent disagreed with this proposition. Notably, one in five respondents (21 percent) supported both of these propositions. Other research has also found that the public often holds contradictory or competing views in relation to criminal justice, leading Doble (2002) to argue that the public can favour multiple sentencing goals and sees them as complementary, not contradictory.

Fitzgerald et al. also found support for rehabilitative ideals, with 82 percent of respondents agreeing that we should ‘spend more money funding effective prison-based education and treatment programs so that people leaving prison do not commit new offences’. They were also optimistic about offenders’ ability to rehabilitate, with 77 percent agreeing that ‘most offenders can go on to lead productive lives with help and hard work’. Notwithstanding this, when asked directly to compare between rehabilitation and community safety, 77 percent regarded community safety as a higher priority.

There was a sense that parole should be regarded as a one-time offer, with 75 percent agreeing that ‘offenders who have violated parole conditions once in the past should never be eligible for parole again’. There was also little support for parole in relation to violent offenders, with only 8 and 15 percent respectively supporting parole for child sex offences or murder, compared with 66 percent for burglary and fraud.

As noted above, non-parole periods in Australia generally range from 50 to 80 percent of the total sentence imposed. This would appear to be out of line with community expectations, with 58 percent of respondents either opposing parole altogether or indicating that prisoners should be required to serve at least 80 percent of their sentence before release.

There were fairly high levels of confidence in parole boards, with 61 percent agreeing that they are in a better position than judges to choose an appropriate release date and 71 percent agreeing that they try to be fair. Unsurprisingly, respondents who supported parole were more likely to agree with these statements. Conversely, while 69 percent of respondents wanted to see victims have more of a say in parole decisions, this was higher among respondents categorised as ‘parole non-supporters’.

Fitzgerald et al. concluded that there is ‘evidently a lack of understanding of the nature and purposes of parole’ (2016: 324). This was also one of the findings of another recent study on parole conducted in Western Australia. In that study, Gately et al. (2017) found that nearly two-thirds of the 38 interview respondents understood that parole involved releasing prisoners back into the community before completing their sentence, but many thought that this involved the sentence being shortened. In addition, although most respondents were aware of the concept of a ‘parole board’, there was limited understanding about how this worked. As set out above, Australia has eight different legislative regimes for parole. One of the concerning findings from Gately et al.’s study is that most participants were not aware that Western Australia has its own parole board.

It appears that some of the questions asked of respondents in this study were highly specific, for example, Gately et al. noted that no participants ‘knew the proportion of offenders in [Western Australia] who receive parole’ (2017: 299) or ‘possessed knowledge about the composition or processes of the [parole] board’ (2017: 301–2). These gaps in specific knowledge are hardly surprising, given the general paucity of knowledge about the criminal justice system identified in previous research. For example, Halstead (2015) found that only 1 percent of NSW respondents correctly identified that less than 10 percent of crimes involve violence.

Just over a third of the respondents in Gately et al.’s study (36 percent) showed positive perceptions of parole; another 36 percent had mixed views and 26 percent expressed negative sentiments. The respondents were subsequently provided with information about parole and parole processes. Although 84 percent of respondents indicated that the information had improved their understanding, only three of the 10 respondents who had held negative views changed their views as a result of this information. This led the authors to highlight the ‘difficulty of changing entrenched opinion cultivated by years of opposing information’ (2017: 306). However, they also found that respondents became more supportive of parole when it was presented in a reintegration framework.

Conclusion and future directions

As noted above, Australia provides an ideal opportunity for examining penal policy and practice, given the variations in law and practice in relation to both sentencing and parole. In addition, there has been a substantial body of research on public opinion on sentencing in recent years and, since 2016, parole. This paper has provided insight into these emerging understandings by presenting an overview of the principal findings in relation to public opinion on sentencing from both top-of-the-head and case-specific research.

Two key themes emerge from this research. Firstly, although respondents typically express dissatisfaction with sentencing (Gately et al., 2017; Lovegrove, 2007; Mackenzie et al., 2012), they tend to impose sentences that are on par with or more lenient than the sentence imposed by the judge when provided with relevant information about sentencing (Lovegrove, 2007; Warner et al., 2009, 2011, 2017a). This is consistent with international research (see e.g. Doob and Roberts, 1983; Gainey and Payne, 2003). It would appear from Warner et al.’s research (2009, 2011, 2017a) that the relationship between information and satisfaction with the sentence imposed may not extend to sexual offences in Australia, although Devilly and Le Grand’s (2015) study suggests that, even in this context, sentencing practices may not be entirely out of alignment with community attitudes. Further research on public opinion on sentencing in sexual offence cases is currently underway (Bartels et al., 2014; Warner, 2014).

The second theme was support for alternatives to prison, especially for non-violent offenders, young offenders, and those with substance abuse issues and/or mental illness (Gelb, 2011a; Mackenzie et al., 2012; Simpson and Butler, 2015; Simpson et al., 2015). Gelb (2011c) and Spiranovic et al. (2012) also found support for rehabilitation for first-time offenders.

The article also presented the key findings from the first two Australian studies on public opinion and parole. These show overlap with the findings on sentencing. For example, both Fitzgerald et al. (2016) and Gately et al. (2017) found low levels of support for parole for violent offenders. Similar to Gelb (2011c), Spiranovic et al. (2012) and Simpson and Butler (2015), Fitzgerald et al. and Gately et al. also found support for rehabilitation. In addition, Fitzgerald et al.’s finding that respondents generally have confidence in parole boards aligns with the finding in Victoria that most respondents thought that courts choose an appropriate sentence most of the time (Gelb, 2011b).

The findings from the sentencing research also point to two future areas of inquiry in parole research. Firstly, it would be of significant interest to disaggregate the data from Fitzgerald et al.’s national study to determine if Roberts, Spiranovic and Indermaur’s (2011) findings – that jurisdictional differences in policy are not related to public attitudes – are replicated in relation to parole. Secondly, the research by Lovegrove and Warner et al. shows that providing members of the public with case-specific information brings their sentences into line with actual judicial practice, at least in respect of offences that do not involve sexual offending (but see Devilly and Le Grand, 2015). Fitzgerald et al.’s ‘parole vignettes’ (2016: 313) will therefore provide a rich source of data to better understand and explore public opinion on parole. Ideally, this would in turn inform policy and help Australia to move away from the punitive trends described in this article.