Terrorism and Parole in the High Court
17.3.2025
In June 2024, the High Court handed down its decision in R v Hatahet [2024] HCA 23 (HCA Judgment) on appeal from the New South Wales Court of Criminal Appeal (CCA) decision in Hatahet v R [2023] NSWCCA 305 (CCA Judgment). The High Court unanimously held that a reduced chance of parole as a result of the operation of s 19ALB of the Crimes Act 1914 (Cth) (Crimes Act) cannot be taken into account in sentencing. Even though terrorist offenders are extremely unlikely to be released on parole, sentencing courts cannot take this into consideration when setting the length of the head sentence and the non-parole period. Section 19ALB requires exceptional circumstances be proven before someone who has committed a terrorism offence can be granted parole. This holding overturned the CCA decision that a sentence served with almost no chance of parole constitutes a more onerous period of imprisonment and, in this case, warranted a reduction in the head sentence.
Central to the High Court’s decision to rule out the consideration of parole prospects in criminal sentencing were three overlapping arguments. The first was that the prospects of parole are too speculative to be taken into consideration during sentencing. In particular, it is inappropriate for the sentencing court to speculate as to how the executive branch will make parole decisions. Secondly, the High Court found it persuasive that it would subvert the legislative intention of s 19ALB of the Crimes Act to take the unlikelihood of parole into account in order to reduce the head sentence. Thirdly, the High Court held that considering s 19ALB was incompatible with the factors that inform judicial sentencing under s 16A of the Crimes Act. Unlike the CCA, the High Court did not accept that reduced parole prospects could be taken into account as contributing to a more onerous period of imprisonment.
In this post, we challenge the arguments accepted in the High Court and compare them with those accepted in the CCA. The ratio of the CCA case was a narrow one, in that it was careful to limit its reasoning to the finding that parole prospects can be considered when they are likely to impact the accused’s conditions of imprisonment. Thus, it was common ground between the CCA and the High Court that parole prospects cannot be considered in isolation during sentencing – this is a point that we do not challenge. Instead, we address the critical question of whether parole prospects can be considered as part of the conditions of imprisonment, and argue that the High Court has unjustifiably limited the discretion of sentencing judges.
Parole
Parole is a form of early release and refers to the conditional release of a prisoner from custody after the expiration of the minimum term of the sentence (the non-parole period). The primary aim of parole is to incentivise rehabilitation by offering prisoners the prospect of early release. This opportunity is believed to foster greater chances of prisoner reform and promote improved overall discipline within correctional facilities (see, eg, NSW Parliamentary Library Briefing Paper).
In sentencing Commonwealth offenders, a non-parole period is generally imposed for sentences over 3 years (Crimes Act ss 19AB, 20(1)(b)). The court can only decline to fix a non-parole period if satisfied it is not appropriate to do so having regard to the nature and circumstances of the offence and the offender’s antecedents, or if the offender is expected to be serving a State or Territory sentence at the end of the federal sentence: s 19AB(3).
There is no prescribed ratio between the non-parole period and parole period; however, the minimum period must reflect the minimum time that justice requires the offender must serve having regard to all the circumstances of the offences: see, eg, Power v The Queen (1974) 131 CLR 623, 628-9 (Barwick CJ, Menzies, Stephen and Mason JJ). The function of directing release on parole resides with the Commonwealth Attorney-General as do decision-making powers such as revoking parole and amending parole conditions (see Judicial Commission commentary).
Background to the case
Mr Hatahet travelled to Syria in 2012 to find and seek the release of his brother-in-law whom he believed was being held by an armed group operating in Syria.[1] After arriving in Syria and meeting his contacts, Mr Hatahet engaged in hostile activities with the Free Syrian Army, a coalition of rebel militias fighting against the al-Assad dictatorship. In 2021, Mr Hatahet was convicted under s 6(1)(b) of the Crimes (Foreign Incursions and Recruitment) Act 1978 (Cth) for engaging in such hostile activities. In 2022, the District Court sentenced him to five years of imprisonment with a non-parole period of three years. Mr Hatahet was classified as an extremely high risk restricted (EHRR) inmate and was held in the High Risk Management Correctional Centre (HRMCC): HCA Judgment at [43].
Mr Hatahet’s parole application was refused in 2023 by the Commonwealth Attorney-General pursuant to s 19ALB of the Crimes Act. Section 19ALB restricts parole for those involved in terrorism-related activities unless exceptional circumstances exist:
(1) Despite any law of the Commonwealth, the Attorney-General must not make a parole order in relation to a person covered by subsection (2) unless the Attorney-General is satisfied that exceptional circumstances exist to justify making a parole order.
(2) This subsection covers the following persons:
(a) a person who has been convicted of a terrorism offence, including a person currently serving a sentence for a terrorism offence;
(b) a person who is subject to a control order within the meaning of Part 5.3 of the Criminal Code (terrorism);
(c) a person who the Attorney-General is satisfied has made statements or carried out activities supporting, or advocating support for, terrorist acts within the meaning of that Part.
It was not in dispute that Mr Hatahet’s conduct would fall within s 19ALB(2)(c).
CCA Judgment
In November 2023, the CCA (Basten AJA; Davies and Cavanagh JJ agreeing) allowed Mr Hatahet’s appeal against his sentence of five years of imprisonment with a non-parole period of three years and resentenced him to imprisonment for four years with a non-parole period of three years.
Basten AJA reasoned that the lack of a realistic chance of release on parole was likely to negatively impact Mr Hatahet’s mental health and so he faced more onerous conditions of imprisonment than the other offenders to whom s 19ALB does not apply. Basten AJA found it reasonable to assume the Attorney-General would likely be satisfied under s 19ALB(2)(c): at [45]-[46]. Subsection (2)(c) covers ‘terrorist acts’ which, read in light of the definition of a ‘terrorist act’ in s 100.1(1) of the Criminal Code, would likely cover foreign incursions or foreign interference such as Mr Hatahet’s conduct.
As to the existence of ‘exceptional circumstances’ in s 19ALB(1), Basten AJA held that they must relate to the offender’s specific situation and general parole purposes are not enough. The Commonwealth Director of Public Prosecutions (DPP) did not suggest that any such circumstances existed.
Basten AJA made a distinction between two scenarios: one is where a sentencing judge considers the likelihood of parole for the purpose of lengthening a sentence; the other is where a sentencing judge considers the likelihood of parole for the purposes of shortening a sentence: at [57]-[58]. In the first scenario, ignoring the prospects of parole is justified because it safeguards the liberty of the offender from probabilistic reasoning. In the second scenario, consideration of the prospects of parole is permitted where ‘the principle upon which release will be refused is fixed by law’ and release is ‘not a mere statistical probability’, as is the case under s 19ALB: at [58].
Basten AJA then held that the likelihood that ‘parole would be refused’ and the fact that ‘the applicant has spent the majority of his sentence in the HRMCC’ meant that the defendant ‘has suffered’ and ‘will continue to suffer more onerous conditions of imprisonment’: at [84]. His Honour emphasised that sentencing judges should not be ‘blinkered’ as to the practical consequences of imposing a non-parole period which has little if any utility: at [52]. In a way, this conclusion is relatively narrow, in that it is not a broad endorsement of factoring parole likelihood into sentencing generally. Parole prospects were considered where there was a factual basis to conclude that it would likely affect the defendant’s conditions of imprisonment.
HCA Judgment
The Court unanimously held that the CCA, in resentencing Mr Hatahet, erred in taking into account the likelihood that parole would be refused.
The plurality (Gordon ACJ, Steward and Gleeson JJ) gave three reasons. First, that it is ‘too speculative for a judge to make a prediction about the probability of a parole order being made’: at [26]. This speculation stemmed from uncertainties about what a parole board, or in this case the Attorney-General, might decide, including the possibility that ‘the legislative scheme, as well as practice and policies, regarding the parole system may validly change from time to time’: at [26]. The plurality also held that such a speculation is inappropriate because the function of considering whether to grant parole is an executive task and irrelevant to any judicial decision-making: at [34].
Secondly, the plurality reasoned that ‘it would subvert the very point of Parliament’s creation of a presumption against parole in s 19ALB to reduce a term of imprisonment on that basis’. The plurality noted that the presumption is ‘Parliament’s response to the specific nature of the threat posed by offenders who are subject to s 19ALB and the greater need to protect the community from those threats’: at [33].
Thirdly, that prospects of parole are not a relevant sentencing factor in s 16A of the Crimes Act. The plurality held that consideration of prospects of parole ‘would lead to outcomes that are inconsistent with a core object of sentencing, namely, the need to ensure that an offender is adequately punished’: at [28]. Relatedly, the expiration of a non-parole period does not create an entitlement to release, and in this case the defendant ‘has not lost his opportunity to be considered for parole’: at [34]. Thus, it was not possible to say that the conditions of Mr Hatahet’s imprisonment were more onerous, because he was still eligible for parole, although it was extremely unlikely to be granted.
Jagot J and Beech-Jones J concurred with the plurality’s views and made additional remarks.
The Three Reasons
Speculation
The High Court’s observation about parole prospects being mere speculation is difficult to accept. Our argument is that under s 19ALB, an offender is extremely unlikely to be granted parole and any uncertainty is not correctly characterised as mere speculation, and is more akin to other, necessarily probabilistic, sentencing considerations which sentencing courts routinely take into account.
In considering Mr Hatahet’s parole prospects, the sentencing court would be potentially speculating about three outcomes. The first is whether Mr Hatahet’s conduct would fall under s 19ALB(2)(c). Admittedly, subsection (2)(c) turns upon the satisfaction of the Attorney-General that a person has been involved in terrorist acts, and that state of satisfaction will not be known at the time of sentencing. This first hurdle did not appear to loom large given the parties accepted that Mr Hatahet’s circumstance would likely satisfy subsection (c), and so it was not expounded upon in great deal in the CCA or High Court. As Jagot J observed, ‘it is not in dispute, that when the respondent was sentenced “there would have been good reason to assume … that the Attorney would be so satisfied in relation to the activities of the [respondent]”’: at [47].
More controversially in this matter, the second area of potential speculation concerns whether the Attorney-General would be satisfied that exceptional circumstances exist to justify making a parole order. Basten AJA recalled that in AH v R [2023] NSWCCA 230, the DPP accepted evidence that none of the 53 applications for parole made by persons subject to s 19ALB resulted in a parole order: at [80]. Indeed, Mr Hatahet had already had his application refused at the time of hearing.
While parole typically arises post-sentencing and depends on the accused’s conduct in prison, s 19ALB changes this by imposing an exceptionally high barrier to parole eligibility based on the nature of the offence. The legislative intention behind s 19ALB was to make it exceptionally hard to get parole. As the plurality observed, ‘[t]he Revised Explanatory Memorandum to the Bill ... describes [s 19ALB] as creating a “presumption against parole” for persons convicted of certain terrorism offences … or who have made statements or carried out activities supporting, or advocating support for, terrorist acts’: at [18]. Thus, parole is no longer merely ‘triggered by events which take place after the completion of an offender's criminal trial’: cf plurality at [37]. The purpose and operation of statute was such as to make the outcome near certain. That there would unlikely be exceptional circumstances reaches a threshold of certainty that is more than mere speculation.
The High Court suggested that legislative schemes cannot be considered because they may change. For example, in Minogue v Victoria (2019) 268 CLR 1, an amendment was made to the Corrections Act 1986 (Vic) many years after Minogue had been originally sentenced which changed the basis upon which he might be granted parole: at [26]. This reasoning is unconvincing as it suggests that the uncertainty of future legislative changes should prevent courts from acknowledging the current, tangible impact of parole on sentencing.
Since the majority was satisfied that parole prospects were too speculative, they did not address a potential third question as to whether the unlikelihood of parole would create more onerous conditions of imprisonment that the court may consider under s 16A(2) of the Crimes Act (as matters “relevant and known to the court”). This question was addressed by Jagot J who doubted whether there was a logical connection between prospects of parole and conditions of imprisonment.
Judges regularly engage with probabilities about the future. In Bugmy v The Queen (1990) 169 CLR 525, Mason CJ and McHugh J held that sentencing factors such as ‘the propensity of the offender to commit violent crimes [and] the likelihood of his re-offending’ are ‘central to a proper discharge of the judicial task’ despite the ‘difficulty of making a forecast of future behaviour so long ahead’: at 533 (emphasis added). In Weininger v R (2003) 212 CLR 629, Gleeson CJ, McHugh, Gummow and Hayne JJ observed that the phrase ‘known to the court’ rather than using words such as ‘proved in evidence’ or some equivalent expression suggests strongly that s 16A of the Crimes Act was not intended to require formal proof of matters before they could be taken into account in sentencing: at 635-8. Courts have also held specifically that findings about such matters as the offender’s prospects of rehabilitation or the risk of re-offending do not always need to be articulated in terms of a standard of proof – ‘assessment by the judge of the risk of a prisoner re-offending is not a feat which requires any refinement of thought process’: eg R v Pickard [1998] VSCA 50 at [4] (Winneke P and Charles JA). The impact of the near-certainty that parole would be refused on conditions of imprisonment is one such relevant matter that can be considered under a less formal evidentiary basis without undermining the integrity of the sentencing process.
The evidence relied upon by the CCA in this case was that the accused’s belief that parole is unlikely will affect their experience of imprisonment including the accused feeling ‘depressed … when he heard about the offenders in the HRMCC being denied parole’: at [77], [80]. While most inmates can improve their prospects of parole by, for example, engaging with rehabilitative programs, those convicted of terrorism offences have even less control over their futures. Thus, not only could it be argued that Mr Hatahet would face more onerous conditions of imprisonment after the expiry of the nominal non-parole period, but he would arguably face more onerous conditions of imprisonment throughout the term. It seems probable that for most if not all offenders, the augmented prospect of serving their entire term of imprisonment in custody would have an adverse impact on their mental wellbeing. A sentence with a nominal non-parole period, as opposed to one where the prospect of parole is real, no doubt weighs more heavily on the offender.
An argument that might be put against us is that, unlike predicting the likelihood of rehabilitation, predicting the likelihood of parole appears to involve speculation as to executive decision making. Arguably, it is permissible to engage with inherent unknowns as to the offender’s future conduct, but impermissible to predict the Attorney-General’s future decisions: Appellant’s Submissions at [47].
However, this distinction between predictions which concern executive decisions and those that do not breaks down upon further analysis. For example, when the court considers the conditions of imprisonment, such as the specific facility where an offender will serve their sentence, they implicitly assume that the executive will not transfer the offender to a different facility or substantially change the conditions within it. As Basten AJA pointed out in the CCA judgment at [52], the sentencing court may take into account that the offender will serve part of their sentence during a lockdown, although it is an executive decision whether to impose lockdowns. In so doing, the court are relying on assumptions about future executive decisions, even though such assumptions are not guaranteed and may be subject to change. This overlap blurs the line between acceptable consideration of future uncertainties and impermissible speculation about executive actions, and we question the legitimacy of the distinction in this case.
Moreover, the CCA’s reasoning is not about considering s 19ALB in isolation without context; but rather about the fact that given the unlikelihood of parole, the accused may face harsher conditions of imprisonment given the specific circumstances of the case. Thus, like other sentencing considerations, the predictions concern the offender’s circumstances, within a context controlled by executive decision-making.
Subverting legislative intention
Two things should be said about the High Court’s conclusion that considering parole prospects subverts Parliament’s intention in creating a presumption against parole under s 19ALB.
The first is that it sits uncomfortably with the High Court’s observations about speculation. If the legislature’s intention is to create a barrier to the release of certain offenders on parole, it is more than a mere ‘guess’ that the executive will not grant parole. A legislative intention of a presumption against parole could surely inform the Court’s conclusion that the Attorney-General would be unlikely to grant parole based on the facts of the case.
Secondly, the plurality again has considered parole prospects as one isolated factor without acknowledging that the CCA was trying to connect this factor with the assumption or argument that it will create more onerous circumstances of imprisonment. Both the CCA and the High Court agreed that prospects of parole are irrelevant on an isolated level: Basten AJA at [50].
We further acknowledge the general principle is that the prospect of securing release on parole or of obtaining remissions is not relevant to the judicial task of sentencing: Hoare v The Queen (1989) 167 CLR 348, 353-4 (Mason CJ, Deane, Dawson Toohey and McHugh JJ). However, the policy against speculation about executive policy regarding remissions emerged from a very different statutory context, one where the liberty of the offender was at risk because sentencing courts could lengthen a sentence with an expectation that the offender would be paroled before it expired: Attorney-General v Morgan (1980) 7 A Crim R 146 at 156 (Jenkinson J; Young CJ and Kaye J agreeing), cited in HCA Judgment at [55] (Jagot J). That consideration is turned on its head under the scheme in s 19ALB, where the offender faces a harsher sentence because the sentencing court closes its eyes to the operation of s 19ALB. As part of the instinctive synthesis in sentencing (Markarian v The Queen (2005) 228 CLR 357, 375 [39] (Gleeson CJ, Gummow, Hayne and Callinan JJ), 377-378 [51] (McHugh J)), any fact relied upon that is ‘adverse to the offender’must be proved beyond reasonable doubt: R v Olbrich (1999) CLR 270 at [27] (Gleeson CJ, Gaudron, Hayne and Callinan JJ) citing R v Storey [1998] 1 VR 359 at 369 (Winneke P, Brooking and Hayne JJA and Southwell AJA). The court requires a high threshold of proof before the prospect of remissions can be taken into account, because the offender’s liberty is at stake. The complaint that parole prospects are speculative holds less weight because those concerns as to liberty do not here apply.
Section 16A
The judiciary’s task is to impose a sentence under s 16A of the Crimes Act. Subsection (1) provides the approach to be taken when sentencing federal offenders: ‘[i]n determining the sentence to be passed … a court must impose a sentence or make an order that is of a severity appropriate in all the circumstances of the offence’. Subsection (2) then provides a non-exhaustive list of factors that can be taken into account on sentence, ‘[i]n addition to any other matters’.
The plurality concluded that ‘the contention that the respondent’s imprisonment was more burdensome because of his reduced prospects of parole is misconceived’: at [34]. They articulated four main reasons. First, the likelihood of getting parole is not a factor judges need to consider when deciding a sentence. Secondly, parole prospects are not part of the ‘circumstances of the offence’ under s 16A(1). Thirdly, considering parole prospects could result in unfair sentencing as it goes against the goal of making sure offenders are properly punished. Fourthly, parole falls entirely within the authority of the executive – ‘there is only ever one sentence imposed by a court … and … the issue of parole is left to the executive branch of government’: at [34]. In our view, none of these reasons are convincing.
In relation to the first two reasons, the plurality was again omitting the critical dispute about whether a sentencing judge can factor in parole prospects where they relate to or affect any of the factors of sentencing in s 16A(2), for instance s 16A(2)(n) regarding the prospect of rehabilitation of the person.
In her Honour’s concurring judgment, Jagot J sought to cut through the ambiguity by explaining how ‘there is no logical connection between a speculated or well-founded lack of potential for an offender to obtain a parole order by reason of s 19ALB(1) and the offender’s conditions of imprisonment’: at [58]. Her Honour criticised the CCA for wrongly assuming that Mr Hatahet’s harsher prison conditions after the non-parole period were due to parole ineligibility, while ignoring that this hardship actually stemmed from his EHRR classification.
Admittedly, the parties offered little specific evidence and did not specify what tougher conditions the accused would or would not face. However, this does not mean that parole prospects are categorically irrelevant to sentencing under s 16A of the Crimes Act. The High Court did not engage with this submission from Mr Hatahet and concluded, without reason and very briefly, that ‘[i]n any event, it should be doubted whether there was a sufficient evidentiary basis for the finding made by Basten AJA that the reduced chance of obtaining release on parole would be likely to adversely affect the mental condition of an offender’, ‘[n]or was there any evidence that the reduced prospect of parole had an effect on issues of deterrence, the prospect here of rehabilitation, or any consequences for the respondent's family or dependants’: at [35].
Regarding the third reason that considering parole prospects is unfair because it undermines proper punishment, it is clear that it would only lead to unfair sentencing if it were to automatically shorten sentences. If, instead, it serves to shorten sentences because it is relevant to the consideration of a legitimate sentencing factor under s 16A, there could not be any problem of unfairness. Section 16A, for example, includes rehabilitation in the list of matters to which the sentencing judge is to have regard: ss 16A(2)(n). Parole prospects can be crucial to rehabilitation in that when parole eligibility seems unlikely, inmates may lose motivation to pursue rehabilitation and that directly impacts the inmate’s reintegration: see eg Mr Hatahet’s submissions summarised by the plurality at [30].
As for the fourth reason – that parole is an executive function – while correct, it doesn't justify disregarding parole prospects in sentencing. Again, the functional distinction between sentencing and parole that the plurality emphasised could not have a bearing on the sentencing judge’s discretion. While it is true that the legislative intention behind s 19ALB is to implement the government’s anti-terrorism policies by creating a presumption against parole, it is not the judiciary’s role to enforce the sentiment behind this provision. Rather, the sentencing judge must, under s 16A of the Crimes Act, provide fair and proportionate sentences in light of deterrence, retributive and rehabilitative objectives.
Even if taking into account prospects of parole under s 19ALB neutralises the policy goal of that section, that is a result of ineffective legislating rather than the application of an impermissible consideration. Indeed, if the CCA’s decision had been upheld, it can be speculated that the Parliament would have passed further amendments to sentencing legislation to make it effective at lengthening time spent in prison. (Conversely, prohibiting a sentencing judge from considering certain facts may merely reduce transparency in the sentencing judge’s reasoning while doing nothing to practically manifest a ‘tough on crime’ outcome.) Of course, it is important that statutes are generally read coherently. But the High Court never justified why the legislative intention behind s 19ALB requires the sentencing court to ignore matters in s 16A(2) that are relevant to the offender. The separation between judicial and executive roles in sentencing and parole must not restrict judicial discretion to consider factors relevant to sentencing, even if they are relevant to executive decisions, when those factors are not purely speculative.
Conclusion
The High Court’s decision in R v Hatahet has, without proper justification, stripped judges of sentencing discretion by creating an artificial and untenable boundary between judicial sentencing and executive parole decisions, and simultaneously blurring this boundary by directing judges to enforce the sentiment behind executive parole policies. It represents a significant misstep in the development of cogent federal criminal justice principles.
[1] See facts on sentence in Hatahet v R [2023] NSWCCA 305 at [17]-[22].
Cherry Tang is a tipstaff to a Judge of the New South Wales Court of Appeal. She holds law degrees from the University of New South Wales and Columbia Law School..
Eden Blair is an associate to a Judge of the New South Wales Court of Appeal. She holds a law degree from the University of Sydney.
Suggested citation: Cherry Tang and Eden Blair, ‘Terrorism and Parole in the High Court’ (17 March 2025) <https://www.auspublaw.org/blog/2025/3/terrorism-and-parole-in-the-high-court>