Stanley v Director of Public Prosecutions (NSW)

Emma-Jane Stanley (the Applicant) pleaded guilty to five counts of knowingly taking part in the supply of a firearm to a person unauthorised to possess it, two counts of knowingly taking part in the supply of a firearm part and three counts of having in possession for supply a shortened firearm. Those were offences contrary to ss 51(1)(a), 51BA(1) and 62(1)(c) of the Firearms Act 1996 (NSW) respectively, attracting maximum penalties of five years’ imprisonment, five years’ imprisonment and 14 years imprisonment if the Crown proceeded by way of indictment. Taken into account, pursuant to s 33 of the Crimes (Sentencing Procedure) Act 1999 (NSW) (the CSP Act), on a “Form 1” were a further 10 offences of knowingly taking part in the supply of a firearm part. In the Local Court, where the Applicant was sentenced, the maximum penalty for each offence was two years’ imprisonment and a fine of $5,500. A further offence of possessing ammunition without authorisation contrary to s 65(3) of the Firearms Act was recorded but, pursuant to s 10A of the CSP Act, no other penalty was imposed.

A statement of agreed facts stated that at some time prior to 18 March 2019, a co-offender, Seth Harvey, stored numerous firearms, firearm parts and ammunition in the Applicant’s home in suburban Dubbo without her knowledge. Some were stored in the back of a vehicle parked in her backyard. Others were stored in a manhole under her house. The Applicant became aware of the firearms, firearm parts and ammunition on 18 March 2019 and agreed that she would hold them until they were supplied to another person known to her. There were discussions involving the Applicant and another co-offender, Brendon Gray, concerning the price, during which the Applicant was told that she would be given $500 of the money for the items if they could be sold for $3,000. On 26 March 2019, the Applicant met with the known person, showed firearms to him and participated in retrieving the ammunition crate from the manhole. The known person gave $6,000 in cash to Mr Gray.

The firearms and firearm parts were identified on the statement of agreed facts, and photographs were annexed. They included single shot and repeating rifles manufactured by Winchester, Browning, Husqvarna, Miroku, Birmingham Small Arms, Omark Sportco and Brno.

The Applicant was arrested on 13 November 2019. She made full admissions to the offences. She said that she had been told that only $3,000 had been paid, of which she took $50. She was granted bail pending sentence.

On 8 December 2020, the Local Court at Dubbo sentenced the Applicant to an aggregate term of imprisonment of three years with a non-parole period of two years, commencing on that day. However, the Applicant filed a Notice of Appeal to the District Court on the same day and was granted conditional bail pending that appeal. The appeal lay pursuant to s 17 of the Crimes (Appeal and Review) Act 2001 (NSW) and was “by way of a rehearing of the evidence given in the original Local Court proceedings”. The appeal was heard by Judge N Williams on 28 May 2021, and the Applicant gave evidence and was cross-examined. Judge Williams reserved judgment, continuing bail, and delivered judgment orally on 17 June 2021, dismissing the appeal and confirming the sentence imposed by the Local Court.

The Applicant presented a strong subjective case. At the time of the appeal, she was aged 38. She had five children, two of whom lived with their father and whom she saw every weekend, and three of whom (a 15-year-old son and four-year-old twins) lived with her. She was single. Her step-father, with whom she maintained a close relationship, was a member of a motorcycle club and her mother had been a heroin addict and physically abusive. After her mother’s death, the Applicant had used ice from around 2013–2015, but not subsequently. She had attended secondary school and had a “significant” employment history with long periods of continuous employment. She had some criminal convictions previously, none especially serious, one of which was served by way of an intensive correction order. She had never previously served a term of full-time imprisonment. A sentencing assessment report described her as having a medium risk of reoffending.

At the forefront of the submissions in the District Court was whether, as the Applicant submitted, the sentence should be served by way of an intensive correction order, or, as the Crown submitted, a sentence of full-time imprisonment was appropriate and the appeal should be dismissed. In neither the written nor the oral submissions of the Crown or the Applicant, however, was s 66(2) of the CSP Act (set out at [18] below) or the specific assessment required by that subsection mentioned. As shall be seen, that sub-section was at the centre of debate in this Court on the application for judicial review.

It is not necessary to summarise the reasons of Judge Williams in any detail. They were lengthy and involved a careful review of the evidence and the submissions that were made before her, consistent with the nature of the appellate task her Honour was performing. Her Honour indicated that she had:

For the purposes of this Summons for judicial review, it suffices to summarise what was said in support of the First Respondent’s submission in this Court that implicitly her Honour had complied with s 66(2). The First Respondent pointed to the fact that in assessing objective seriousness, her Honour had said:

In determining the length of the sentence to be imposed, her Honour took into account in the sentencing exercise “particularly general deterrence which must loom large particularly specific deterrence and of course community safety and denunciation”: at p 28. Her Honour concluded that the three-year sentence imposed by the Local Court was an appropriate term.

Her Honour then concluded (at pp 28–29):

There was debate during the hearing in this Court as to whether “pose” as appears in the transcript was accurate, or whether her Honour had in fact said (or intended to say) “posed”.

Following the dismissal of her appeal to the District Court, the Applicant commenced serving the sentence imposed.

The Applicant’s Summons in this Court’s supervisory jurisdiction was filed on 28 September 2021.

Pursuant to that Summons, the Applicant seeks relief in the nature of certiorari quashing the decision of Judge Williams, together with an order that the proceedings be returned to the District Court for the Applicant’s appeal from the Local Court to be determined in accordance with law.

Judge Williams was exercising the jurisdiction conferred on her by s 11 of the Crimes (Appeal and Review) Act. By s 17 of that Act, and as noted above at [5], such an appeal is by way of “a rehearing of the evidence”. This has been interpreted in a series of cases as involving a hearing de novo: see, for example, Lunney v Director of Public Prosecutions (NSW) [2021] NSWCA 186 at [20]; cf McNab v Director of Public Prosecutions (NSW) [2021] NSWCA 298 at [25] in relation to appeals pursuant to s 18 of the Crimes (Appeal and Review) Act. Accordingly, her Honour considered the question of sentencing afresh but dismissed the appeal such that the sentence imposed by the Local Court stood.

Relevantly for present purposes, her Honour was asked to and considered whether or not the sentence should be served in the community by way of an intensive correction order, the judge having a discretion to so order pursuant to s 7(1) of the CSP Act which relevantly provides that “[a] court that has sentenced an offender to imprisonment in respect of 1 or more offences may make an intensive correction order directing that the sentence or sentences be served by way of intensive correction in the community.” The power to make an intensive correction order has been considered in a number of decisions of this Court and the Court of Criminal Appeal: see, for example, R v Fangaloka [2019] NSWCCA 173; Casella v R [2019] NSWCCA 201; Blanch v R [2019] NSWCCA 304; Mandranis v R [2012] NSWCCA 97; Mourtada v R [2021] NSWCCA 211.

Section 66 of the CSP Act provides that:

The essence of the application for judicial review in the present case was that, whilst Judge Williams undoubtedly considered whether an intensive correction order was appropriate in the circumstances of the case, her Honour did not, in the course of her consideration, conduct the assessment contemplated by s 66(2) of the CSP Act, and her failure to do so amounted to jurisdictional error.

The Applicant called in aid in support of that submission this Court’s decision in Wany v Director of Public Prosecutions (NSW) (2020) 103 NSWLR 620; [2020] NSWCA 318 at [67] (Wany). The correctness of that decision, insofar as it was held that the “failure to consider the matter identified in s 66(2)” was a jurisdictional error, was challenged by the Director of Public Prosecutions (NSW), who is the First Respondent to these proceedings. It was for that reason that I convened a bench of five judges to hear this application.

Subsequent to the hearing of this application, this Court has delivered its judgment in Quinn v Director of Public Prosecutions (Cth) [2021] NSWCA 294 (Quinn) which held that McCallum JA (with whom Simpson AJA relevantly agreed) was wrong to hold at [67]–[68] of Wany (albeit perhaps somewhat tentatively) that the failure to consider the matter identified in s 66(2) was a jurisdictional error. Simpson AJA, who was also a member of the Bench in Quinn, withdrew her support for agreement with that conclusion in Wany: Quinn at [191].

For the reasons given in Quinn, I agree that Wany was wrongly decided in this respect and should not be followed. In this respect, I also agree with what Leeming JA has said in his separate reasons in relation to supplementary submissions filed on behalf of the Applicant following the handing down of judgment in Quinn.

In argument in this Court, Mr Quilter, with whom Ms O’Neill appeared for the Applicant, placed emphasis on the need, under s 66(2), to conduct an assessment “as to whether making [an intensive correction] order or serving the sentence by way of full-time detention is more likely to address the offender’s risk of reoffending” (the assessment process). This was arguably a narrower proposition than that which had been accepted in Wany and was said to require in effect a two-stage process, with the assessment process first to be undertaken and then the result of that assessment to be taken into account “when considering community safety”.

The argument also depended upon a conclusion that the judge had not in fact undertaken the assessment process contemplated by s 66(2). Ms Richardson SC and Ms Gleeson for the Director were unable to point to any explicit part of her Honour’s reasons which evidenced either the assessment process being performed or the results of that assessment process being articulated, although it was submitted that it could be inferred that her Honour had undertaken the assessment process in the course of her consideration of the appropriateness or otherwise of making an intensive correction order, cf Chalhoub v R [2021] NSWCCA 69.

Because of my ultimate conclusion, there is no need in the present circumstances to make a definitive finding as to whether or not Judge Williams in fact engaged in the assessment process although I am inclined to agree with Beech-Jones JA’s assessment that her Honour did not do so.

These reasons proceed on the basis that no such assessment process was in fact undertaken. Had it been undertaken, absent some denial of procedural fairness, it could not have been impugned for jurisdictional error even if the Court took the view that the judge’s assessment was wrong. There was no suggestion of a denial of procedural fairness.

On the assumption, therefore, that the assessment process was not in fact undertaken, it is necessary to consider whether or not that posited failure constituted jurisdictional error.

The starting point is, of course, that there is no right of appeal, even by way of leave, from a decision of a District Court judge hearing an appeal in a criminal matter from the Local Court. So much is reinforced by s 176 of the District Court Act 1973 (NSW): see, for example, Jamal v Director of Public Prosecutions (NSW) [2019] NSWCA 121 at [8]; see also and more generally, Gibson v Commissioner of Police (NSW) (2020) 102 NSWLR 900; [2020] NSWCA 160 at [20]–[21].

Section 176 of the District Court Act and its predecessor, s 146 of the Justices Act 1902 (NSW), have long been identified as privative clauses: see, for example, Director of Public Prosecutions (NSW) v Emanuel (2009) 193 A Crim R 552; [2009] NSWCA 42 at [45]; and Hoffenberg v District Court of New South Wales [2010] NSWCA 142 at [4].

Privative clauses such as s 176 are, by definition, designed to limit the scope for judicial review, multiple appeals and, in some cases, any appeal: see, for example, the Jurisdiction of Courts (Cross-vesting) Act 1987 (NSW) s 13; Summary Offences Act 1988 (NSW) s 27(2). Such clauses play an important role in the legal system, furthering the value of finality and, subject to the Constitution, effect must be given to their underlying purpose in accordance with s 33 of the Interpretation Act 1987 (NSW) and common law principles of statutory interpretation as qualified by the presumption that “the legislature does not intend to deprive the citizen of access to the courts, other than to the extent expressly stated or necessarily to be implied”: Public Service Association (SA) v Federated Clerks’ Union (SA Branch) (1991) 173 CLR 132 at 160; [1991] HCA 33.

In Anderson v Judges of District Court (NSW) (1992) 27 NSWLR 701 at 718 (Anderson), Kirby P (with whom Meagher and Sheller JJA agreed) observed in relation to s 146 of the Justices Act, expressed in identical terms to s 176 of the District Court Act, that:

Where a privative clause such as s 176 of the District Court Act is engaged and no appeal (or no further appeal) is provided for from a decision of the District Court, the only other means of review available to a party seeking to challenge that decision requires the identification of jurisdictional error. The important legacy of Kirk v Industrial Court of New South Wales (2010) 239 CLR 531; [2010] HCA 1 (Kirk) lies in the High Court’s constitutional insistence that judicial review for jurisdictional error may not be validly circumscribed or circumvented by privative clauses. In that case, the relevant clause was s 179(1) of the Industrial Relations Act 1996 (NSW).

“Jurisdictional error” is, thus, a concept of fundamental importance to the work of this Court. It is the only avenue available, for example, to a person such as the Applicant who has been convicted of an offence in the Local Court of New South Wales, was unsuccessful in his or her appeal to the District Court of New South Wales and who wishes to seek further review. Thus, perhaps somewhat paradoxically, a not insubstantial portion of the subject matter of the work of the New South Wales Court of Appeal (as opposed to the Court of Criminal Appeal) is criminal in nature, coming to this Court through the invocation of its supervisory jurisdiction. Proceedings for judicial review of this kind have been held, at least for certain purposes, not to change their character from civil to criminal despite the underlying subject matter: Klewer v Director of Public Prosecutions (NSW) (No 2) (2020) 101 NSWLR 864; [2020] NSWCA 69 at [38]–[44], [107]–[113] where it was held that proceedings for judicial review were not “criminal proceedings” within the meaning of s 8(9) of the Vexatious Proceedings Act 2008 (NSW); cf Chief Executive Officer of Customs v Labrador Liquor Wholesale Pty Ltd (2003) 216 CLR 161; [2003] HCA 49 at [121].

Before Kirk, there was Craig (Craig v South Australia (1995) 184 CLR 163; [1995] HCA 58). Decided just over 25 years ago, Craig ensured that, in Australian law, “jurisdictional error” remained conceptually discrete from other errors of law: at 177–180. This was by way of contrast with the position in England where, in Anisminic Ltd v Foreign Compensation Commission [1969] 2 AC 147, there had been an elision of jurisdictional error and error of law: see also O’Reilly v Mackman [1983] 2 AC 237 at 278; R v Hull University Visitor, Ex parte Page [1993] AC 682 at 701–702.

Whilst jurisdictional error may be described as an error of law, not every error of law will be “jurisdictional” or amount to jurisdictional error. This is not a matter of semantics. It is of cardinal importance. An error of law made within jurisdiction is not a jurisdictional error: Gibson v Director of Public Prosecutions (NSW) (No 2) [2021] NSWCA 218 at [31]–[33], [48]–[53], [64] (Gibson) citing Public Service Association and Professional Officers' Association Amalgamated Union of New South Wales v New South Wales (2014) 242 IR 338; [2014] NSWCA 116; Shade Systems Pty Ltd v Probuild Constructions (Aust) Pty Ltd (No 2) (2016) 95 NSWLR 157; [2016] NSWCA 379.

As Basten JA (with whom Bathurst CJ and Beazley P agreed) observed in Wang v Farkas (2014) 85 NSWLR 390; [2014] NSWCA 29 at [42], echoing the observations of Kirby P in Anderson set out above:

The continuing vitality of the concept of jurisdictional error in Australian law as confirmed in Craig has not been matched by clarity in its definition or elucidation and, as the High Court was to point out some 15 years after Craig, eight different categories of jurisdictional error had been identified to that point in time: see Kirk at [71], citing M Aronson, “Jurisdictional Error Without the Tears” in M Groves and H P Lee, Australian Administrative Law: Fundamentals, Principles and Doctrines (2007, Cambridge University Press) at 335–336.

The lack of clarity as to what constitutes jurisdictional error was partly explained by the Kirk majority’s frank acknowledgement that the content of the concept of jurisdictional error has evolved over time and is not necessarily characterised or marked by logical coherence: at [62]–[63]. This was in part attributed to the varied usages and meanings of the word “jurisdiction”. Also acknowledged at [64] was the realist view of Professor Jaffe, in “Judicial Review: Constitutional and Jurisdictional Fact” (1957) 70 Harvard Law Review 953 at 963, that denominating some questions as "jurisdictional":

A “rigid taxonomy of jurisdictional error” was eschewed by the majority in Kirk (at [73]), which was also at pains to point out that the three examples of jurisdictional error instanced in Craig were “just that – examples”.

A further dimension and level of complexity has arguably been added to the jurisprudence in this area by Hossain v Minister for Immigration and Border Protection (2018) 264 CLR 123; [2018] HCA 34 at [31] (Hossain) where, building on the joint judgment of Gageler and Keane JJ in Wei v Minister for Immigration and Border Protection (2015) 257 CLR 22; [2015] HCA 51 at [23]–[32], the concept of materiality was identified as relevant to a consideration of whether a jurisdictional error will result in invalidity; see also at [66] per Edelman J. In other words, it may now be necessary to speak of a tri-partite division between jurisdictional error the commission of which results in the invalidity of a decision or order; non-material jurisdictional error which does not; and error of law within jurisdiction (non-jurisdictional error). In the present case, no question of materiality arose. The debate was simply whether or not the error identified and relied upon by the Applicant was jurisdictional error.

Consistent with the notion that the class of what constitutes “jurisdictional error” continues to evolve and is not closed, in their sixth edition of Judicial Review of Administrative Action and Government Liability (2017, Lawbook Co) at [1.140] Professors Aronson, Groves and Weeks proffered a revised catalogue of 10 categories of jurisdictional error, with categories nine and 10 having been added to the list of eight categories compiled by Professor Aronson and referred to in Kirk at [71]. The revised list is as follows:

The scope for “terminological entanglement” was noticed by Kiefel CJ, Gageler and Keane JJ in Hossain at [17] and cases have continued to multiply as to what is and what is not jurisdictional error, with different views being taken at the intermediate appellate level: for recent examples in this Court, contrast Gibson at [31]–[33], [42] and Franklin v Director of Public Prosecutions (NSW) [2021] NSWCA 83; Wany and Quinn. See also Blissett v Director of Public Prosecutions (NSW) [2021] NSWCA 253 at [34].

Such differences highlight the “difficulty of drawing a bright line between jurisdictional error and error in the exercise of jurisdiction” (Re Refugee Review Tribunal; Ex parte Aala (2000) 204 CLR 82; [2000] HCA 57 at [163]) but, as cases such as the present illustrate, and as Leeming JA explained in Quinn, the drawing of the line may have immense practical significance. In that context, non-compliance with mandatory language in a statute is only the starting point of the analysis as to whether or not jurisdictional error has been committed.

As the plurality pointed out in Hossain at [27], statutory construction is critical to a consideration of the question; see also at [72] per Edelman J. As with any exercise in statutory construction, close attention must be given to the statute as a whole and statutory context: Project Blue Sky Inc v Australian Broadcasting Authority (1998) 194 CLR 355; [1998] HCA 28 at [69].

Particular techniques of legislative drafting may assist in demarcating where the line is to be drawn between jurisdictional and non-jurisdictional error through the use of clear formulae such as “a relevant order is not invalidated by a failure to comply with this section” or some variation thereon. Examples of the use of such formulae are furnished by ss 5(4), 17I(2), 17J(4), 25F(8), 53A(5) and 100B(2C) of the CSP Act and non-compliance even in the face of unqualified mandatory language such as “must ensure”, “must cause”, “must indicate and … must record”, and “must indicate … and make a written record” employed in these sections will not constitute jurisdictional error.

Other examples of this useful drafting technique may also be seen in ss 18H(2) and 18I(2) of the Drug Court Act 1998 (NSW); ss 9E(3), 9F(2) and 13F(2) of the Graffiti Control Act 2008 (NSW); ss 5(6) and 16F of the Child Protection (Offenders Prohibition Orders) Act 2004 (NSW); s 48M(2) of the Children (Criminal Proceedings) Act 1987 (NSW); and ss 80(5) and 89A(5) of the Fines Act 1996 (NSW). In this context, in New South Wales at least, there appears to have been something of a salutary shift from the position expressed by Basten JA in Area Concrete Pumping Pty Ltd v Childs (2012) 223 IR 86; [2012] NSWCA 208 at [81], where his Honour observed that it was “rarely the case that a parliament specifies with precision the intended consequences of a failure to comply with a provision, except negatively by means of a privative clause”.

On the issue of legislative drafting, it should also be observed that infraction of a statutory prohibition which expressly conditions the Court’s jurisdiction on the existence of a particular objective fact will more obviously amount to jurisdictional error: see, for example, Reysson Pty Ltd v Minister Administering the Environmental Planning and Assessment Act 1979 [2020] NSWCA 281 at [55]; Muswellbrook Shire Council v Hunter Valley Energy Coal Pty Ltd [2019] NSWCA 216; (2019) 372 ALR 695. An example is arguably supplied by s 68(1) of the CSP Act which provides that “an intensive correction order must not be made in respect of a single offence if the duration of the term of imprisonment imposed for the offence exceeds 2 years”; see also s 67(1). It is not necessary to decide, in the present case, whether the Director’s concessions that non-compliance with ss 4B(1) and 17F(1) of the CSP Act would amount to jurisdictional error were correctly made.

The understanding of concepts such as jurisdictional error, whose varied nature and manifestations may make definition difficult, can sometimes usefully be illuminated by a statement of what the concept is not, that is to say, by negative definition. In this context, in an important passage in Kirk, the majority held, citing Craig at 179–180, that demonstrable error on the part of an inferior court “‘entrusted with authority to identify, formulate and determine’ relevant issues, relevant questions, and what is and what is not relevant evidence” would “not ordinarily … constitute jurisdictional error”: at [67].

The composite expression “identify, formulate and determine” a “relevant issue” seems to me aptly to capture what a sentencing judge is entrusted to do under s 66(2) of the CSP Act, namely to identify a relevant issue, being the offender’s risk of reoffending according to whether or not he or she is incarcerated or permitted to serve the sentence in the community, to formulate the competing matters bearing upon that risk and to determine the issue. This is an exercise in which the commission of demonstrable error would “not ordinarily constitute jurisdictional error” according to the Kirk majority.

One question which arises is what the majority meant by the phrase “not ordinarily”. In my view, although it is not entirely clear, their Honours may have here been allowing for the possibility that the conduct of the exercise entrusted to the Court, being an exercise within jurisdiction, could nonetheless amount to jurisdictional error if, for example, it was tainted by a denial of procedural fairness. That is not this case. Any other interpretation of the expression “not ordinarily” in the passage from Kirk would be to erode further the already limited certainty as to the meaning of the concept of “jurisdictional error”.

On the assumption that Judge Williams did not assess the Applicant’s risk of reoffending according to whether her sentence was to be served in prison or in the community under an intensive correction order, in accordance with s 66(2) of the CSP Act, such a demonstrable error was not jurisdictional in nature. This conclusion has the perhaps awkward but in my view unavoidable consequence that, had the Applicant been convicted on indictment and the same assumed error been committed in sentencing, an appeal would lie, with the grant of leave, to the Court of Criminal Appeal: Criminal Appeal Act 1912 (NSW) s 5(1)(c); cf, for example, Karout v R [2019] NSWCCA 253.

On the other hand, had the Applicant been convicted on indictment, it is quite likely that she would have been exposed to a far greater penalty than that able to be (and in fact) imposed in the Local Court. Indeed, given that some of the offences to which the Applicant pleaded guilty carried a maximum penalty of between five and 14 years’ imprisonment (see [1] above), she may well not have been eligible for an intensive correction order at all on that alternative scenario: see s 68(1) and (2) of the CSP Act, restricting the cases where an intensive correction order may be made to sentences not exceeding two years in respect of a single offence and not exceeding three years in the case of an aggregate sentence.

An important consequence of the more limited scope for judicial review on the basis of jurisdictional error in the context of s 66 of the CSP Act, when compared to the exercise of a right of appeal with leave pursuant to s 5(1)(c) of the Criminal Appeal Act, is that grounds that may attract appellate intervention in relation to consideration of an intensive correction order by reference to House v The King (1936) 55 CLR 499; [1936] HCA 40 will not necessarily be available on an application for judicial review; see also s 101A of the CSP Act. This is because, in the context of an appeal, non-jurisdictional as well as jurisdictional errors of law may be relied upon. Practitioners seeking to bring applications for judicial review to the Court of Appeal from decisions of the District Court must therefore not make the mistake of assuming that a ground upon which an offender has succeeded in the Court of Criminal Appeal will necessarily be available on an application for judicial review to the Court of Appeal.

Earlier in these reasons, reference was made to ss 5(4), 17I(2), 17J(4), 25F(8), 53A(5) and 100B(2C) of the CSP Act and the use of formulae in these sections of the Act which made plain that non-compliance with an express statutory provision would not result in the invalidity of the order or sentence in question. The presence of such provisions gives rise to a potential expressio unius argument, namely that the absence of an equivalent formulation in respect of non-compliance with s 66(2) of the CSP Act should, by implication, result in invalidity. Such an argument should not, however, be accepted. As I said in Australian Rail Track Corporation Limited v Dollisson [2020] NSWCA 58 at [47] (Dollisson):

As with the Accident Compensation Act 1985 (Vic) under consideration in Dollisson, the CSP Act is a frequently amended Act and s 66(2) is a recent inclusion effected at a different time to those sections to which I have drawn attention in [45] above. Whilst expressio unius reasoning may be a “valuable servant”, it can be a “dangerous master”: Colquhoun v Brooks (1888) 21 QBD 52 at 65; see also Daniels Corporation International Pty Ltd v Australian Competition and Consumer Commission (2002) 213 CLR 543; [2002] HCA 49 at [34]; Houssein v Under Secretary, Department of Industrial Relations and Technology (NSW) (1982) 148 CLR 88 at 94; [1982] HCA 2.

Reliance on expressio unius reasoning cannot, in my opinion, provide a foundation for a conclusion which leads to and results in the serious (and potentially adverse) consequences of invalidity which the construction of the CSP Act propounded by the Applicant would entail. These are pointed out by Basten JA at [127]–[137] below. His Honour’s reasons, as with those of Leeming JA in Quinn, are a textbook illustration of the need to consider the question at hand by construing a statutory provision in the full context of the Act in which it appears, by reference to the nature of the task in which the court is engaged and with a keen eye to the full consequences and ramifications of invalidity, not only for the immediate decision in question but for the class and volume of decisions likely to be affected by the consequences of non-compliance. A full appreciation of this last matter will often supply a significant pointer to the question of legislative intent as to the consequences of non-compliance in those cases where the consequences have not been expressly addressed in the parliamentary drafting.

It would also be a most peculiar outcome if non-compliance with the prohibition in s 5(1) of the CSP Act, namely that “[a] court must not sentence an offender to imprisonment unless it is satisfied, having considered all possible alternatives, that no penalty other than imprisonment is appropriate”, did not result in invalidity (which is what s 5(4) of the Act expressly provides) but that non-compliance with the assessment process contemplated by s 66(2) did have that consequence. McCallum JA has suggested, at [170] below, that s 5(4) of the CSP Act, which provides that “a sentence of imprisonment is not invalidated by a failure to comply with this section”, should be construed as relating only to non-compliance with s 5(2) and not s 5(1) of the CSP Act. That construction, with respect, finds no support in the unqualified language of s 5(4) of the CSP Act nor, so far as I can ascertain, is the departure from the plain meaning of s 5(4) warranted by considerations of legislative history or particular context of the kind that led Leeming JA, Emmett AJA and me recently (and exceptionally) to depart from the literal meaning of s 11(1) of the Federal Courts (State Jurisdiction) Act 1999 (NSW) in Sydney Seaplanes Pty Ltd v Page [2021] NSWCA 204; (2021) 393 ALR 485.

The nature of what was essentially propounded as a jurisdictional fact, namely the undertaking of what I have earlier described as “the assessment process” embedded in s 66(2), also points strongly against a conclusion that non-performance of it amounted to jurisdictional error. In Plaintiff M70/2011 v Minister for Immigration and Citizenship (2011) 244 CLR 144; [2011] HCA 32 at [57], French CJ observed, citing Australian Heritage Commission v Mount Isa Mines Ltd (1997) 187 CLR 297 at 303–304; [1997] HCA 10, that:

Whilst I agree generally with Basten JA’s reasons which I have had the benefit of reviewing and which lead to the same conclusion that I have reached, I would reserve my position in relation to his Honour’s discussion and consideration of s 43 of the CSP Act. That is an important provision in respect of which the Court did not have the benefit of argument. My conclusion that the assumed error of which the Applicant complains was not a jurisdictional error is not dependent upon a consideration of s 43 of the CSP Act.

Although the principal thrust of the Applicant’s argument was that undertaking the assessment contemplated by s 66(2) was a condition of the exercise of jurisdiction — an argument which I have rejected — to the extent it was put on the alternative basis that there was jurisdictional error based upon a misconception by the judge of her function – an argument that has found favour with McCallum JA – I would also reject that argument.

The function being performed by the District Court was hearing and determining the Applicant’s appeal against sentence. As noted at [16] above, that function was conferred by s 11 of the Crimes (Appeal and Review) Act.

Judge Williams was acutely aware of the Applicant’s contention that her sentence should be served in the community, subject to an intensive correction order: see, for example, the passages from her Honour’s judgment set out at [8], [9] and [11] above. I do not doubt that her Honour gave that matter earnest consideration. That her Honour may not have expressly or even impliedly made an assessment of how the Applicant’s risk of reoffending would be impacted by reference to where and how the sentence was served (an exercise she was not asked by the parties to undertake and in respect of which no submissions were made to her Honour) did not, in my view, mean that she misunderstood the nature of her function. Nor did it mean that community safety was not taken into account as the paramount consideration (see s 66(1) of the CSP Act). Judge Williams’ reasons make it explicit that that matter was considered together with other factors identified in s 3A of the CSP Act, as required by s 66(3). It was certainly not something that was “swept aside”: cf, McCallum JA at [176]. If there was any error as to how community safety was taken into account, it was an error within jurisdiction, not amenable to judicial review by this Court in its supervisory jurisdiction.

Her Honour’s function was dictated by the jurisdiction she was exercising, namely the hearing of an appeal de novo as to the sentence imposed on the Applicant in the Local Court, and a consideration of whether the sentence should have been made the subject of an intensive correction order. Her Honour performed and did not misconceive this function, let alone fundamentally so. To hold that a judge has “fundamentally misconceived” his or her function is a serious conclusion which should, in my view, only be made in a clear case. A misconception of function is not the simple corollary of a failure to comply with a statutory provision; were it otherwise, every such failure would translate into a jurisdictional error.

The summons for judicial review should be dismissed. Each party should bear her own costs of the proceedings.

BASTEN JA: The applicant, Emma-Jane Stanley, seeks a review in this Court of an aggregate sentence imposed on her in the District Court following an appeal to that Court from the Local Court. There is no right of appeal from the judgment of the District Court; further, there is a privative clause in s 176 of the District Court Act 1973 (NSW), the effect of which is to limit this Court’s supervisory jurisdiction to review for jurisdictional error.

None of the above is contentious. However, recent decisions in this Court reveal different approaches to the scope of the Court’s jurisdiction in such cases. A five judge bench has been convened to resolve those differences.

The subject matter of the application was a sentence of 3 years imprisonment, with a non-parole period of 2 years. The sentence, originally imposed in the Local Court, followed pleas of guilty entered by the applicant in relation to 10 firearms offences, including seven offences of knowingly taking part in the supply of a firearm and three offences of possession for supply of a shortened firearm. All of the offences were serious, the latter three offences each carrying a maximum penalty of 14 years imprisonment. There were ameliorating circumstances in relation to the applicant’s role which warranted them being dealt with in the Local Court, rather than on indictment. Nevertheless, the applicant received the maximum aggregate sentence available in that Court.

On appeal to the District Court, Judge N Williams confirmed the sentences. As stated by the solicitor appearing for the applicant in the District Court, there was no issue but that a custodial sentence should be imposed; the issue was whether the sentence should be served by way of an intensive correction order (ICO). Judge Williams held that the sentence should be served by way of full-time custody. The appeal was dismissed.

Given the disparity in approaches which has been revealed in recent cases, the appropriate course is to address the question of jurisdiction as a matter of principle, before turning to the decisions of this Court.

It is not uncommon for applications for judicial review to be formulated in terms which reflect the table of errors found, for example, in the Administrative Decisions (Judicial Review) Act 1977 (Cth), s 5. As the tabulation was taken from the general law, it is understandable that the statutory language has been widely adopted beyond the scope of the Commonwealth Act. Thus, grounds frequently assert a failure to have regard to a relevant consideration, or taking into account an irrelevant consideration. There are, however, three pitfalls in this approach.

First, not all formulations of grounds distinguish jurisdictional error from other errors of law. Secondly, there is an important distinction to be drawn between the scope of authority conferred on an administrative decision-maker and the jurisdiction of a court. In particular, a court will be authorised to determine questions of law, a function which is not usually conferred on an administrative decision-maker, at least where there is a firm separation between executive and judicial power. Thirdly, it follows that statements of principle in cases and texts dealing with administrative decisions must be carefully scrutinised for relevance before being translated to review of judicial decisions.

More fundamentally, and putting to one side the specific issues concerning non-statutory executive power, most administrative decision-making is authorised by statute; it follows that the limit of the decision-maker’s authority will depend upon questions of statutory construction. The same is not true of a court, including a statutory court, such as the District Court of New South Wales. While every State court has limits on its jurisdiction and powers, fundamental aspects of the workings of a court are not to be found in statute. Principles of procedural fairness, the obligation to give reasons, and indeed the scope of the supervisory jurisdiction of this Court, are not to be found in statute, nor will they be identified by a process of statutory construction.

Particularly is that so in relation to the function of sentencing, a function once compared by Jordan CJ to that of a jury in a civil trial assessing damages where the assessment of value is not the subject of any accepted measure. [1] In Veen v The Queen [No 2], [2] Mason CJ, Brennan, Dawson and Toohey JJ stated:

Parliament can, of course, impose obligations and constraints on the sentencing process. However, in construing statutory provisions relevant to sentencing, it is necessary to bear in mind the fundamental nature of the task of sentencing and the manner in which it is carried out by a court.

The importance of having regard to the institutional nature of the decision-maker was expressly addressed in Craig v South Australia [5] where the Court rejected the identification of forms of error by Lord Reid in Anisminic Ltd v Foreign Compensation Commission [6] as “not intended to refer to a court of law.”

Craig is commonly cited for its statement of principle as to what constitutes jurisdictional error, although the statements were to some extent qualified in Kirk v Industrial Court of New South Wales. [7] What should be noted, however, was that Craig involved an appeal from an order of the Full Court of the Supreme Court of South Australia setting aside a decision of a judge of the District Court to stay proceedings until the accused had representation. The Full Court held that the judge had made factual and legal errors in applying the decision of the High Court in Dietrich v The Queen. [8] The High Court held that the South Australian Full Court had been wrong to intervene. The High Court concluded: [9]

The authoritative statements of the nature of jurisdiction and jurisdictional error found in the joint reasons of Kiefel CJ, Gageler and Keane JJ in Hossain v Minister for Immigration and Border Protection [10] are carefully confined to review of administrative decision-making. That is not to say they are not valuable in the context of judicial review of judicial decision-making, but the principles require adaptation.

Before turning to the terms of the Crimes (Sentencing Procedure) Act 1999 (NSW) (Sentencing Procedure Act) it is important to note that the ultimate question is whether, in giving direction to the sentencing court as to how to address particular matters, Parliament intended to invalidate any sentence passed where there was non-compliance with the statutory direction. That the answer to that question turns on the proper construction of the statute is established by the reasoning of the High Court in Project Blue Sky Inc v Australian Broadcasting Authority. [11] In considering the nature of a statutory obligation the joint reasons quoted the following passage from Francis Bennion’s Statutory Interpretation: A Code: [12]

Although identified in the context of administrative decision-making, the following statements of principle are equally applicable to the construction of the Sentencing Procedure Act. Thus Brennan CJ (dissenting as to the outcome, but not the approach) stated:

The majority (McHugh, Gummow, Kirby and Hayne JJ) spoke in similar terms in relation to the general principle:

The circumstances and reasoning in Tasker v Fullwood are instructive. Tasker concerned an application by a lessee of premises for a liquor licence. The Liquor Act 1912 (NSW), s 24(1A), required that where there was a person with an interest in the business, “the applicant shall produce to the licensing court an agreement between such person and the applicant”, to the effect that the applicant would have unfettered control of the conduct of the business. No such agreement was produced, but a licence was issued. The terms of the statutory provision clearly imposed an obligation; nevertheless, the Court (Hope, Glass and Samuels JJA) stated: [15]

Finally, it is essential for a court which primarily deals with civil appeals by way of rehearing under s 75A of the Supreme Court Act 1970 (NSW), or a Court of Criminal Appeal dealing with sentence appeals under s 5(1)(c) of the Criminal Appeal Act 1912 (NSW), a jurisdiction subject to the constraints explained in House v The King, [16] to bear in mind that the supervisory jurisdiction is of a different character. The instinctive response is to assess the merit of the decision under appeal; that exercise plays no part in the Court’s exercise of its supervisory jurisdiction, particularly where restricted by a privative clause to jurisdictional error.

As noted above, the principal question in the District Court proceeding was whether the applicant should be required to serve her sentence of imprisonment in detention, or in the community under an intensive correction order. The power to make an intensive correction order is to be found in s 7 of the Sentencing Procedure Act which provides as follows:

Part 5, referred to in s 7(4), comprises ss 64-73B, divided into four divisions. In Div 1, s 64 provides that the Part applies “in circumstances in which a court is considering, or has made, an intensive correction order.” Nothing seems to turn on the past tense, “or has made”, which may be intended to pick up the requirements of Div 3 as to the term of an order and as to the commencement of an order, namely “on the date on which it is made”: s 71(1).

Relevantly for present purposes, Div 2, with the heading “Restrictions on power to make intensive correction orders”, contains s 66 upon which the argument turned. Section 66 reads as follows:

The applicant relied upon three propositions, namely, first, that s 66(2) imposed a duty to undertake a balancing exercise as to which alternative form of sentence was more likely to address the risk of reoffending; secondly, that the sentencing judge did not undertake that exercise, so that, thirdly, the order dismissing the appeal was a nullity.

The question whether the judge failed to undertake the balancing exercise may be put to one side for the moment. The underlying principle for statutory construction must be addressed in two parts, namely (i) what is the nature of the obligation imposed by s 66(2), and (ii) did Parliament intend that a failure to comply with that obligation would invalidate the sentencing process?

This issue must be addressed by applying the reasoning in Project Blue Sky and Tasker v Fullwood. The same conclusion should be reached in the present case as in Tasker, though the reasons differ and the reasons which were powerful in Tasker have not been set out above. The relevant reasons in this case are as follows.

By way of introduction the nature of provisions conferring or constraining jurisdiction fall into different categories. Without seeking to be comprehensive, these include (i) factors independent of the exercise of judicial discretion; (ii) factors resulting from a prior exercise of judicial discretion; (iii) fundamental principles governing the administration of criminal justice; (iv) statutory constraints applying to sentencing generally, and (v) statutory constraints applying to the exercise of a particular power. Examples are, as to (i), setting the maximum sentence the Local Court can impose; as to (ii), permitting an ICO where the judge has decided to impose a sentence of imprisonment for a single offence of not more than 2 years; as to (iii), principles of procedural fairness; and as to (iv), s 21A setting out factors to be taken into account in sentencing. The present case falls within (v), which may be that least likely to include pre-conditions to the valid exercise of a power.

It is necessary first to identify the form of the obligation. Although it is true that the relevant words (“is to assess”) confer a function on the sentencing court, it is a secondary function, and the consequences of a particular assessment are not identified. It is secondary in the sense that the assessment is for the purpose of considering “community safety”; community safety is said to be “the paramount consideration” when deciding whether or not to make an ICO: s 66(1). That is, the obligation is not directed to the outcome of the sentencing exercise, but to a step along the way.

Section 66(2) must be read in context. The sentencing court is also obliged to consider the provisions of s 3A and “any relevant common law sentencing principles”. These principles involve evaluative judgments, the consequences of which are not spelled out. Thus, the section does not say that an ICO must not be imposed unless it is more likely to address the offender’s risk of re-offending; nor that it must be imposed if the balance favours the likelihood of a lower risk of re-offending. To give effect to the criteria specified in s 66(1) and (2) without regard to other relevant matters would contradict the mandate in subs (3). Finally, there is no express indication that a sentence is invalid unless the assessment is undertaken. Should such a conclusion be inferred? The following considerations militate against any such inference.

The conceptual framework provides the relevant statutory context. The central focus is something identified as “community safety”. That concept is not defined, but may be taken to include protection of the community from the offender, being a specific purpose identified in Veen v The Queen [No 2] and in s 3A(c) of the Sentencing Procedure Act. That section, introduced at the same time as s 66, states:

Although the purposes of sentencing do not fall into any simple dichotomy, the purposes stated as (a) adequate punishment, (f) denouncing conduct, and (g) recognition of harm done, may be seen to be primarily backward-looking; the other provisions may be understood as primarily forward-looking. There is clearly overlap between (b) personal and general deterrence, (c) protecting the community from the offender and (d) promoting rehabilitation of the offender. The forward-looking purposes would all be relevant to “community safety”, but in different ways and at different points in time. At least in the short term, community safety may best be protected by a period of full-time custody, followed by a period of supervised liberty. On the other hand, where there is confidence in an existing process of rehabilitation in the community which is already underway, the longer term likelihood of re-offending may be diminished by not sending the person to prison. Section 66(2) says nothing about how such temporal considerations should feed into the assessment. In short, the nature of the assessment is one which cannot be clearly defined, involves evaluative assessments for which there may be no factual basis, and which may not lead to any useful conclusion. It is implausible to infer that the legislature intended to characterise such an obligation as a condition of validity of the sentencing process whenever an ICO is being considered.

The provision should be assessed in the context of the statute taken as a whole. The Sentencing Procedure Act is replete with directions to sentencing courts. Those which might be thought to constitute the clearest examples of limitations on jurisdiction are prohibitions. In relation to ICOs there are express prohibitions in relation to a number of categories of offending, including offences involving the discharge of a firearm: s 67(1)(f). (That prohibition was not engaged in the present case as the applicant’s offending did not involve the discharge of a firearm; nevertheless, the category of offence will be a highly material consideration.)

Section 5, dealing with penalties of imprisonment generally, contains a prohibition and a mandatory obligation, but also contains an express statement that non-compliance with either or both will not invalidate the sentence.

There are a number of provisions which impose express obligations in mandatory terms which are accompanied by a statement that an exercise of the power is not invalidated by non-compliance: see, eg, s 17I, s 17J, s 25F, s 53A, s 100B, s 100P. Generally these sections use the language of what the court “must” do.

There is another category of provisions which states what a court “is to” do. These include s 21A which identifies matters which the court “is to take into account” in determining the appropriate sentence: s 21A(1). The factors include both aggregating and mitigating factors and total more than 30: most are routinely ignored because they are not relevant in particular circumstances or because, even if potentially relevant, there is no evidence before the court in relation to them. There are various ways in which the obligation identified in s 21A could be read down so as to exclude unknown or irrelevant factors, or factors upon which neither party relies. In such a case, the test would generally rely upon a reasonable state of satisfaction of the trial judge, not on the opinion of a reviewing court. So far as this Court is aware, there has never been a case in which review for jurisdictional error has relied upon a failure to comply with s 21A.

There are numerous sections which use language similar to that of s 21A, such as requiring that the court “is to have regard to” specified matters: see, eg, s 9(2), s 10(3), s 25D(6), s 30K(2). Of relevance to ICOs, s 69 provides:

There are some 40 statements in the Act that the court “is to” undertake a particular function or exercise. If only because the consequences of each entailing invalidity would be so inconvenient, if not destructive of the orderly operation of the criminal justice system, it is implausible that the Parliament expected invalidity to flow as a result of non-compliance with all or even some of these provisions. Although some are accompanied by a saving clause (that non-compliance does not entail invalidity) there is no clear inference that the absence of such a clause should lead to a different effect. [17] In many cases the obligation is procedural and no issue would necessarily arise as to the validity or otherwise of any subsequent court order. In short, a provision saying what the sentencing court “is to” do does not by itself demonstrate an intention to impose a condition of validity.

The Sentencing Procedure Act contains a general provision dealing with the consequences of serious sentencing defects.

The scope of the reopening power in s 43(2) was considered by the High Court in Achurch v The Queen. [18] The precise operation of the provision is not presently relevant; however, its purpose is significant. As the joint reasons of French CJ, Crennan, Kiefel and Bell JJ recognised, “[a]bsent specific statutory authority, the powers of courts to reopen their proceedings and to vary their orders is constrained by the principle of finality.” [19] That principle is qualified by the power of a court to correct an error before orders are formally recorded, [20] and by the “slip rule”. [21] The joint reasons stated:

The joint reasons gave as examples of errors engaging s 43 a penalty which exceeds the maximum penalty prescribed for the offence and the imposition of a penalty which is beyond power “because some precondition for its imposition is not satisfied”. [22] The reasons continued:

Gageler J, agreeing in the outcome, said that s 43 is only engaged if “the order imposing the penalty is in its terms an order that the court could not have made in the criminal proceedings.” [23]

It follows that, if the judge had imposed an ICO in circumstances where the power was not available (perhaps because the sentence exceeded the limit fixed by s 68), such an error could have been corrected under s 43 by “reopening” the proceeding and amending the order. The circumstances in which the section is engaged describe a typical case of jurisdictional error, though may not cover all such errors. The applicant submitted that while the judge had a discretionary power either to grant or refuse an ICO, her decision was legally ineffective because she failed to undertake a specific exercise which might or might not have affected the outcome. Such an error may not fall within the terms of s 43 and therefore, at least by inference, not be capable of correction by the District Court. Whilst it may be accepted that there are such errors (for example, where the offender has not been accorded procedural fairness, or there is a reasonable apprehension of bias) the Court should be slow to construe a statutory requirement as jurisdictional so as to avoid the serious adverse consequences for the administration of justice noted above. If the error is not able to be corrected, there being no right of appeal, that is a consequence of the legislature declining to provide two levels of appeal for less serious criminal offences.

Returning to the terms of s 66, it is necessary to have regard to the purpose or object underlying the provision so as to adopt a construction which would promote that purpose or object rather than a construction which would not: Interpretation Act 1987 (NSW), s 33. The purpose or object will often be derived from the terms of the provision, read in context, but may be assisted by having regard to extrinsic material, including the second reading speech of the Minister: Interpretation Act, s 34. Statutory context has been addressed.

The parties placed reliance upon a passage in the Minister’s second reading speech in the following terms: [24]

These observations suggest that the Government was concerned with two matters when s 66 was introduced into the Sentencing Procedure Act. First, there was a concern that persons were released into the community without supervision; secondly, there was a concern that “short prison sentences” were not effective at reducing reoffending.

It is not entirely clear that these remarks were all addressed to intensive correction orders. The statements as to effective sentencing options appear to have derived from the NSW Law Reform Commission 2013 report on Sentencing. That report identified short sentences of imprisonment as “normally defined as six months or less”. [25] It noted that, in 2012, 34% of all offenders in the Local Court sentenced to terms of imprisonment received a fixed term sentence of six months or less. [26] There were no statistics or observations directed to the effectiveness of longer terms of imprisonment.

The Commission identified certain benefits of home detention and ICOs over full-time imprisonment, including “that they allow an offender to retain employment and remain in contact with family networks while serving the sentence.” [27] Their daily cost per offender to the State was 10% of the cost of full-time custody; they avoided “any potential contaminating effects arising from offenders, and particularly first time offenders, being in prison with other offenders”; they allowed offenders to retain public or community housing and allowed a benefit to the community through community service work. [28] It was also said that home detention produced “relatively low recidivism rates” but that no data were available on rates of recidivism after completion of an ICO.

In an accompanying report providing statistics, [29] the Commission noted, with respect to home detention:

No similar figures were available for ICOs.

A report produced by the Bureau of Crime Statistics and Research in December 2013 [30] compared the risk of reoffending of those who received an ICO with those who received periodic detention or suspended sentences with supervision: there was no comparison with groups of offenders who received full-time detention. At least by implication, the need to match profiles of offenders suggested that a similar study with those who were sentenced to full-time detention might be difficult. No such study was identified in the evidence.

Further, the sentencing court considering whether to make an ICO “is to have regard to … the contents of any assessment report obtained in relation to the offender”: s 69(1)(a). The obligation is stated in somewhat different terms in s 17D:

The unqualified language of s 17D(1) is in fact qualified by subs (2), suggesting that “must not” is used to indicate a default position. Similarly, the effect of s 69(1) is to require the court to have regard to the contents of an assessment report, if there is one.

These reports are addressed in cl 12A of the Crimes (Sentencing Procedure) Regulation 2017 (NSW):

In the present case, an assessment report was provided which referred to the offender’s risk of reoffending. However, it did not, nor was it required to, attempt a relative assessment of risk depending on whether the offender were to serve a sentence in full-time detention or in the community under an ICO. The fact that the court is generally to have regard to an assessment report, but that such a report is not required to provide a basis for the assessment required by s 66(2), suggests that there is no unqualified obligation imposed by s 66(2). Further, the direction in s 17D(1) is not engaged where the sentencing judge is not satisfied that such an assessment can be undertaken, or is otherwise satisfied that an ICO should be imposed.

In many provisions where departure from an obligation is permitted, the court is required to give reasons for non-compliance. It is arguable that the absence of such an obligation in s 66 indicates that no departure is permitted. For the reasons set out above that intention is implausible. The alternative inference is that Parliament did not expect that departure from the obligation necessarily called for an explanation.

When a regime of standard non-parole periods was introduced into the Sentencing Procedure Act in 2002, s 54B read: [31]

In discussing the operation of these provisions in Muldrock v The Queen [32] the High Court commenced its analysis with the following observations:

The question in Muldrock was how the statement in s 54B(2) that the court “is to set” the standard non-parole period should be construed. The High Court stated:

The significance of this reasoning is that, whilst in terms setting an obligation to achieve a particular outcome, s 54B was construed as requiring the judge “to identify fully the facts, matters and circumstances which the judge concludes bear upon the judgment that is reached about the appropriate sentence to be imposed.” [46] Quite apart from the provision of s 54B(5), upon which the High Court did not place reliance, it is clear that the nature of the obligation was properly ascertained by reference to all matters relevant to the sentencing exercise and not by the imposition of a straightjacket dependent purely upon an assessment of objective seriousness.

A close analogy may be drawn with the construction of s 66. Not only must it be read as a whole, but it must be read in the legislative context provided by the Act taken as a whole, including ss 3A and 21A, and against the background of the sentencing function as understood under the general law. (There are several provisions of the Sentencing Procedure Act, including s 21A, which demonstrate that the Act is not a sentencing code.)

Following Muldrock, s 54B was recast so that s 54B(2) now provides:

A second indication of the use of language of obligation in a sense which is directive rather than fixing the limits of power or jurisdiction, is to be found in s 44 of the Sentencing Procedure Act in relation to the setting of non-parole periods. Section 44(1) provides:

Despite the unambiguous grammatical construction of this provision, it has been held that it does not require the sentencing judge first to set a non-parole period and then proceed to consider the length of the overall sentence. As Davies J (with the agreement of Hoeben CJ at CL and Hidden J) stated in Anjoul v R, [47] s 44 “is a procedural provision directed specifically to the establishment of a ratio between the non-parole period and the balance of the term.” To read it as requiring that the non-parole period be identified first before considering the overall term would, Davies J noted, “be inconsistent with provisions in the Sentencing Procedure Act such as s 6 (home detention), s 7 (intensive correction orders) and s 12 (suspended sentences).” Of these only s 7 remains, but it provides that an ICO is a manner of serving a sentence of imprisonment and requires that the Court “has sentenced an offender to imprisonment” but then provides that “the court is not to set a non-parole period for the sentence”: see s 7(1) and (2) respectively.

It may also be observed that s 44 contains four statements in apparently mandatory language, namely subs (1), (2), (2B) and (2C). Section 44(3) states that non-compliance with the last three does not invalidate the sentence; there is no similar express protection given with respect to subs (1). Yet the result is clearly the same.

The joint reasons in Project Blue Sky at [91] noted that the consequences of construing a provision as a pre-condition to the engagement of a power may tell against a legislative intention of invalidity resulting from non-compliance. As Dixon J concisely observed in Parisienne Basket Shoes Pty Ltd v Whyte, [48] in relation to whether a statutory limitation constituted a jurisdictional fact:

Also in Parisienne Basket Shoes, Latham CJ noted: [49]

Parisienne Basket Shoes concerned the Local Court: a similar approach has been adopted in a civil case with respect to the District Court in Pelechowski v Registrar, Court of Appeal (NSW). [50] Mr Pelechowski was found guilty of contempt of court for disobedience in relation to an asset preservation order made by the District Court. Because the District Court had no power to make such an order, the failure to comply with it could not be a contempt of court. The joint reasons in Pelechowski referred with apparent approval to a statement by McHugh JA in this Court in Attorney-General (NSW) v Mayas Pty Ltd [51] to the following effect: [52]

This reasoning pre-dated the general adoption of “jurisdictional error” as a label for invalidity. It appears to suggest that if the District Court imposed a sentence of imprisonment and refused an ICO, both orders being within power, the failure to consider a mandatory consideration would not result in the kind of invalidity to which McHugh JA referred. While the approach in Mayas has never been rejected and is inherently attractive, it may be doubted whether it accurately reflects the current jurisprudence of the High Court with respect to jurisdictional error.

In 2002, in Minister for Immigration and Multicultural Affairs v Bhardwaj, [53] the High Court held that an administrative tribunal, which had made a decision in circumstances which denied procedural fairness to the applicant, had the power to disregard its first decision and reconsider the application. Gaudron and Gummow JJ, in reasons with which McHugh J agreed, Hayne J writing in similar terms, stated:

Gleeson CJ in Bhardwaj approached the issue in the following manner:

No statutory intention precluding the Tribunal from revisiting a flawed decision was found in Bhardwaj.

The consequences of invalidity were adverted to in the course of oral argument in this matter, but were not fully explored. In fact, the consequences of invalidity resulting from non-compliance with s 66(2) are potentially dramatic and disruptive of the orderly administration of justice. A number of factors supporting this conclusion may be briefly stated.

First, the vast bulk of cases in which an ICO will (or should be) considered will be sentences imposed in the Local Court. The possibility that the speedy and result-oriented manner of dealing with a vast number of guilty pleas will, almost inevitably, lead to a significant proportion of cases in which the terms of s 66(2) are not addressed in the reasons of the magistrate and may well not have been appreciated by those appearing. It may also be assumed that there will be individuals appearing unrepresented who may wish to obtain an ICO and who will not know the legal consequence of invalidity flowing from non-compliance with a provision which is not expressed as a condition of obtaining an order.

If the failure of an inferior court, such as the Local Court, to comply with s 66 will result in a nullity which any person can disregard, a person refused an ICO will be unlawfully imprisoned and a person granted an ICO, who was not on bail, may be returned to custody. To conclude that Parliament intended such results is untenable.

In the present case, it was accepted that if the decision of the District Court was invalid, the applicant remained subject to the sentence of imprisonment imposed in the Local Court. However, as she had obtained appeal bail pending determination of her appeal to the District Court, it was also assumed that that bail would continue, so that she would be entitled to liberty pending the uncompleted appeal to the District Court. However, that conclusion involved at least two assumptions which were not explored. The first was that a failure to comply with s 66 in considering whether to grant an ICO rendered the sentence of imprisonment (and not merely the refusal to grant an ICO) a nullity. Secondly, if it did, there was a further assumption that the bail granted by the magistrate, which required her to attend at the hearing of her appeal and was continued by the judge when judgment was reserved on 28 May 2021, was not spent when she attended for sentence on 17 June 2021, or if it were spent, would somehow revive on the setting aside of the sentencing judgment. All of these assumptions may be correct, but they are not self-evidently so. Indeed, the mere fact that the conclusion that the sentence was invalid must lead to a range of complex and difficult questions upon which the liberty of the offender will depend demonstrates the unlikelihood that Parliament intended invalidity as the consequence of non-compliance with s 66(2).

Each of the matters set out above demonstrates that s 66 does not in its terms impose a constraint on the powers or jurisdiction of a sentencing court; it is better described as a direction to the sentencing court as to, first, the fact that considerations which may promote the safety of the community are to be given special (paramount) weight and, second, that one factor to be included in the instinctive synthesis is an assessment of whether the risk of the offender re‑offending would be more likely under an ICO or by requiring full-time detention. It is a reminder to the court that giving paramount effect to community safety does not require incarceration.

Importantly, there is no staged process of consideration, contrary to Markarian, nor is there a requirement to disregard any other relevant consideration; rather there is an obligation in equally emphatic terms to have regard to all other relevant factors. To construe subs (2) in isolation from the rest of s 66 would be an error; to construe s 66 in isolation from other requirements of the Sentencing Procedure Act would be to contradict its own subs (3); to treat s 66 as imposing a mandatory limit on the powers of the sentencing court to fix a valid sentence would be to contradict the approach adopted in cases such as Muldrock and Anjoul.

It follows that whether and how the sentencing judge complied with s 66 is not an issue which can engage the supervisory jurisdiction of this Court. The summons should be dismissed. No party having sought costs, no order as to costs should be made.

LEEMING JA: I do not repeat the background, which is set out elsewhere. I proceed on the basis, favourably to Ms Stanley, that the sentencing judge did not in fact assess as required by s 66(2) of the Crimes (Sentencing Procedure) Act 1999 (NSW) whether making an intensive correction order or serving the sentence by way of full-time detention was more likely to address Ms Stanley’s risk of reoffending. To be fair, her Honour was at no stage asked in terms to do so, although her Honour expressly and repeatedly stated that she had regard to community safety in more general terms and it is evident that community safety was a significant matter informing the exercise of the sentencing discretion. Ms Stanley had not brought the cache of rifles and ammunition onto her property in suburban Dubbo, but she permitted them to be stored there, for eight days, in a vehicle parked in her backyard and in a manhole under her house, where she lived with young children. Three of those weapons, including a .32-20 Winchester 1892 lever action repeating rifle, had been shortened through removing the stock (these led to the charges under s 62(1)(c) of the Firearms Act). Two single shot rifles, one bolt action repeating rifle, one lever action repeating rifle and a double barrel shotgun were stored in a condition where they could be fired. In the period during which the firearms and ammunition were stored at her home, the Applicant was personally involved in negotiations for their sale and in delivering them to the purchaser who paid $6,000.

In oral and written submissions in this Court, attention was drawn to the “gravity” of the error for which the Applicant contended. It was put thus orally:

In further submissions supplied after the hearing, the Applicant contended that the “gravity of the error of failing to undertake the specific assessment process mandated by” s 66(2) supported the conclusion that there was jurisdictional error.

I am prepared to accept for the sake of resolving these submissions the Applicant’s argument that a sufficiently “grave” error by a court might be capable of amounting to jurisdictional error, although I am conscious that doing so runs the risk of reasoning by reference to extreme cases. But I do not accept that the present case comes anywhere near such a “grave” case.

The assessment required by s 66(2) is one component of “community safety” which itself is one of a large number of matters required by the statute to be taken into account in the course of considering whether to make an intensive correction order. While community safety is expressed to be the “paramount” consideration, the aspect of community safety reflected in s 66(2) may be of greater or lesser significance in any particular case.

In the present case, there is good reason to think that the assessment required by s 66(2) was of relatively limited significance, although I am conscious that the District Court, which saw and heard Ms Stanley give evidence, was better placed than this Court to undertake that assessment. The risk to community safety from offending of the type charged was palpably real. The gravamen of her evidence was that the firearms had been placed in her house by her cousin, but she nonetheless participated in their sale rather than calling the police because “I didn’t want to get my, my cousin into trouble and get him put into gaol”. It was open to proceed on the basis that the offending was not driven by financial reward, but by familial and social pressure. If s 66(2) were applied in its terms, then there would be nil chance of re-offending for the period during which Ms Stanley was in full-time detention. It might also be thought that serving the sentence in this way might cause those who sought to use her home to store firearms to refrain from doing so in the future, and if they did, Ms Stanley would be well placed to resist participating in the storing and sale. On the other hand, if the sentence were served by way of an intensive correction order, it might be thought that the risk of Ms Stanley being prevailed upon to participate in similar conduct in the future would not be reduced to the same extent, in circumstances where previous 9 month and 12 month intensive correction orders imposed in November 2015 (and served concurrently) for driving with a high range PCA and while disqualified had not prevented further driving offences occurring in February 2017 shortly after their completion. In addition to the above, the assessment would involve considering the likely effect of an extended period in full-time detention, and the accompanying dislocation from employment and social structures upon the risk of reoffending upon release.

Further, the discretion to make or refuse to make an intensive correction order will be informed by other considerations, both concerning community safety and also concerning other sentencing purposes. The point of the previous paragraph is that it tends to suggest that in the particular facts of this offending, the narrowly framed assessment required by s 66(2) was apt to have been far from the forefront of the considerations bearing upon whether or not to impose a sentence of full-time imprisonment or one served by way of an intensive correction order. That is consistent with what in fact occurred in the Local Court and in the District Court, where despite extensive submissions being made in support of and against an intensive correction order, the assessment required by s 66(2) seems at no stage to have been mentioned. Instead, submissions were couched in more general language directed to community safety, to which her Honour responded. “[W]hen a court is invited to make a discretionary decision, to which many factors may be relevant, it is incumbent on parties who contend on appeal that attention was not given to particular matters to demonstrate that the primary judge’s attention was drawn to those matters, at least unless they are fundamental and obvious”: Macedonian Orthodox Community Church St Petka Inc v Eminence Petar (2008) 237 CLR 66; [2008] HCA 42 at [120]. The fact that neither s 66(2) nor the assessment it requires was mentioned suggests that neither side regarded it as especially significant in the circumstances of the Applicant’s offending.

Whether or not a court’s failure to have regard to something to which statute requires it to have regard amounts to jurisdictional error turns on the statute. Ms Stanley’s written submissions, which had been prepared prior to the delivery of judgment in Quinn v Commonwealth Director of Public Prosecutions [2021] NSWCA 294, were understandably framed in terms of what had been held in Wany v Director of Public Prosecutions (NSW) (2020) 103 NSWLR 620; [2020] NSWCA 318 concerning the effect of failing to have regard to a matter mandated by statute. Making the assumption that s 66(2) was not complied with, that does not amount to jurisdictional error, for the reasons stated by this Court in Quinn. There is no reason to elevate the obligation in s 66(2) into one which, if breached, results in jurisdictional error on the part of the District Court, such that its order is a nullity in law and Ms Stanley’s appeal remains undetermined.

Ms Stanley made further submissions following delivery of judgment in Quinn, in accordance with this Court’s directions. It was put that a failure to perform the assessment required by s 66(2) was to misconceive the District Court’s “function”, which was defined as ordering or not ordering an intensive correction order. It was said that that submission was not directly undermined by the reasoning in Quinn, which focussed on whether disregard of a matter that statute required to be taken into account as a condition of jurisdiction was jurisdictional error. It was contended that this conclusion “would represent an expression of the gravity of the error of failing to undertake the specific assessment process mandated by the Act”. It was said that this would not lead to a multitude of jurisdictional errors, insofar as s 66(2) involved a “process” as opposed to a mandatory consideration, and was supported by the emphatic language of s 66(1) which refers to community safety as the “paramount” consideration. It was also noted that Quinn concerned an ex tempore decision, in contrast with the reserved judgment of the District Court in the present case.

Ms Stanley’s further written submissions were advanced attractively and succinctly, as had been her oral submissions. But I cannot accept most of them. I agree that the submission now advanced is distinct from that determined in Quinn. But I do not agree with the narrow characterisation of the “function” which is said to have been “misconceived”.

First, the function being performed by the District Court was hearing and determining Ms Stanley’s appeal against sentence. That function was conferred on the District Court by ss 11 and 17 of the Crimes (Appeal and Review) Act 2001 (NSW), to hear and determine the appeal against the sentence imposed by the Local Court.

Secondly, counsel for the applicant correctly conceded that even when the issue on appeal reduces to how a sentence of imprisonment for three years is to be served – by intensive correction or by full-time imprisonment – the issue is not binary, because of the obligation to impose additional conditions accompanying an intensive correction order unless the court is satisfied of special circumstances. It is true that it is substantially binary. But the fact that one aspect of the process of determining the appropriate sentence is substantially binary does not make that aspect a separate function for the purpose of determining whether the function was so fundamentally misconceived as to amount to jurisdictional error.

Thirdly, the assessment required by s 66(2) was not “discrete”, as the Applicant submitted, in any legally meaningful sense. The assessment required by s 66(2) was a component which contributed to the assessment of “community safety”, which was itself but one consideration, albeit the paramount consideration, amongst a wide range of considerations bearing upon the sentencing discretion.

The question is whether the orders resolving Ms Stanley’s appeal disclose jurisdictional error. That question is not answered by considering whether one aspect of the process adhered to statute, and then asserting that because it did not, there was a misconception of the District Court’s function within the meaning of that term in the context of jurisdictional error.

Further, I do not regard labelling the assessment required by s 66(2) as a “process” impacts upon the application of principle in any material way. Nor do I accept that the error of failing to perform the assessment under s 66(2) was especially grave. For the reasons summarised above, the assessment was not central to the sentencing discretion, and appears not to have been relied on by either side at two hearings.

Finally, nothing turns on the fact that in Quinn the District Court gave ex tempore reasons while in the present case its judgment was reserved.

Subject to the above, I agree generally with what has been said by Bell P, Basten JA and Beech-Jones JA concerning whether the failure to comply with s 66(2) amounted to jurisdictional error.

Courts which sentence offenders to terms of imprisonment perform a very serious function. So do courts which uphold decisions to refuse protection visas, knowing that the consequence is detention followed by deportation to a place where the applicant claims he or she will suffer harm or death. So do courts which make decisions about the custody of children with a parent, or under the protection of the State. So do courts making adoption orders, or in the exercise of their parens patriae jurisdiction. So do courts exercising civil jurisdiction which make findings of serious criminality such as historical childhood sexual abuse. All these decisions change people’s lives directly and immediately. The function of ordering that a sentence not exceeding three years be served by way of intensive correction as opposed to full-time detention is far from the only function which is of the utmost seriousness. It is true that the principles governing the way in which this Court supervises the exercise of power are informed by the nature of the power. But it does not follow that any different approach applies to the functions of courts exercising criminal jurisdiction resulting in sentences to full-time imprisonment. To the contrary, the principles governing the review of executive and judicial power for jurisdictional error, which are to be regarded as grounded in the Commonwealth Constitution, are universal.

I agree with the orders proposed by Bell P.

McCALLUM JA: I have had the benefit of reading, in draft, the judgments of Bell P and Basten, Leeming and Beech-Jones JJA. I respectfully disagree with their Honours’ conclusions as to the determination of the application. The comprehensiveness with which the issues are addressed in their judgments permits me to state my reasons for dissent briefly.

As Beech-Jones JA has noted, the first issue raised by the application is whether the sentencing judge failed to make the assessment required by s 66(2) of the Crimes (Sentencing Procedure) Act 1999 (NSW). His Honour has concluded that, in determining the applicant’s appeal, the judge did not make the required assessment. I agree, for the reasons his Honour has given. That issue is not determined in the other judgments, a matter to which I will return.

The second issue is whether it follows that the decision entailed jurisdictional error. Contrary to the conclusion reached in all of the other judgments, I have concluded that it does. That is because, in my respectful opinion, the failure to grapple with the issue identified in s 66(2) as it arose in this case reveals a fundamental misconception on the part of the judge of the function her Honour was required to discharge. This is essentially the proposition I accepted (somewhat tentatively, as Bell P correctly notes) in Wany v Director of Public Prosecutions (NSW) (2020) 103 NSWLR 620; [2020] NSWCA 318. The submissions in that case conceived the error differently (as a failure to have regard to a mandatory consideration). However, as has been noted, the authorities in this field are not to be understood to provide a rigid taxonomy. While my brief analysis of the issue in Wany has been tested and found wanting by what has been said in Quinn v Director of Public Prosecutions (Cth) [2021] NSWCA 294 and in the other judgments in the present case, I remain of the view that, in a case in which the issue of the manner in which a sentence of imprisonment is to be served properly arises (I will explain what I mean by that), a failure to apply the section that mandates a process for assessing the consideration designated by parliament as “paramount” in that determination reveals a fundamental misapprehension of the nature of the statutory function. In my view, that is a misconception of authority which goes to the heart of the sentencing function and is qualitatively different from an error of law within jurisdiction.

The other judgments have explained at length the important propositions and principles against which the issue of jurisdictional error is to be determined. They include the requirement to give effect to the purpose of the privative clause and not to “emasculate” it or “denude” it of its intended effect; the emergence of a concept of materiality in determining whether any error (if established) will result in invalidity; the cardinal importance of distinguishing between an error of law made within jurisdiction and a jurisdictional error; and the significance in that context of the distinction between the scope of authority conferred on an administrative body and the jurisdiction of a court, the latter having authority to decide questions of law incorrectly.

I do not take issue with those principles. However, as affirmed in Kirk v Industrial Court of New South Wales (2010) 239 CLR 531; [2010] HCA 1 at [63], the notion of jurisdiction takes its colour from its context. The sentencing function is unique. It is the institutional mechanism for maintaining the rule of law by force of the power to impose punishment, including the sternest punishment available, imprisonment in a correctional centre. It is difficult to conceive of a more solemn power. The assistance to be obtained from comparing a statutory command as to the exercise of that power with a statutory command that governs an application by a lessee of premises for a liquor licence (cf Basten JA at [81]) or the circumstances in which an owners’ corporation can initiate legal action (cf Quinn at [117]) is, with respect, limited. The task of discerning Parliament’s intention as to the consequence of non-compliance with a statutory command governing the sentencing function must be undertaken having regard to the solemn nature of the sentencing power and its institutional importance in the governance of the State.

Basten JA has explained that, whereas the limit of an administrative decision-maker’s authority is a matter of statutory construction, the same is not true for courts: “fundamental aspects of the workings of a court are not to be found in statute” (see [72]). But it does not follow that there is no limit on a court’s authority to decide. What does follow is that the limit of that authority might not be found in the literal words of the statute. In my view, the limit of the court’s authority to commit a person to imprisonment in a correctional centre must be informed by the institutional character of that authority. Perhaps that is simply another way of saying that, in construing a statutory provision according to its text and context, the solemnity of the power to imprison is an important aspect of its context.

One factor that has been viewed by the other judges as militating against the conclusion that the failure to comply with the command of s 66(2) was intended by Parliament to invalidate the sentence imposed is the fact that the outcome of that single command may not be determinative in the sentencing process. I accept that a failure to have regard to a single relevant factor in determining an appropriate sentence might not of itself amount to jurisdictional error. However, depending on the factor and the context in which it arises for consideration, it might. Bell P has noted the qualification in Kirk at [67] (drawn from Craig v South Australia (1995) 184 CLR 163; [1995] HCA 58) where the Court noted:

Contrary to what Bell P has tentatively suggested, I do not think that qualification was to allow for the possibility that the conduct of an exercise within jurisdiction could nonetheless amount to jurisdictional error if, for example, it was tainted by a denial of procedural fairness. Denial of procedural fairness is not a species of error in the identification, formulation and determination of what is relevant; it is a different kind of error altogether. In my view, the qualification acknowledged in Craig and endorsed in Kirk allows for the possibility that demonstrable error on the part of an inferior court in identifying, formulating and determining relevant issues, relevant questions, and what is and what is not relevant evidence will not ordinarily but might constitute jurisdictional error. I would read the qualifying words “not ordinarily” as a concession to the impossibility of finding the holy grail of certainty in this field.

I said I would return to the approach taken by others of determining the question of jurisdiction without first determining whether the judge in fact failed to make the required assessment. For my part, I have difficulty seeing how the possibility that a particular error is jurisdictional can be excluded by considering the words of the statute alone. I accept that the question of jurisdiction is one of statutory construction. But I do not see how it can be concluded, in reliance only on the language and drafting techniques of the statute, that any departure from the statutory command, no matter how demonstrable and how egregious, will be within power. Testing a proposed construction by reduction to an example of the absurd is a recognised technique. Consideration of the particular example alleged in the case serves the same purpose.

This point can be illustrated by reference to s 5 of the Crimes (Sentencing Procedure) Act. That section provides:

Bell P has construed s 5(4) to mean that failure to comply with s 5(1) does not result in invalidity. On that premise, his Honour reasons that it would be a peculiar outcome if non-compliance with the assessment process required by s 66(2) did have that consequence. I agree that would be a peculiar outcome but I disagree with the premise. His Honour’s construction of s 5 would hold that, if a judge was satisfied that a fine would be an appropriate penalty but nonetheless proceeded to sentence the offender to a term of imprisonment, the sentence would nonetheless be valid so that, if imposed in the District Court following an appeal under the Crimes (Appeal and Review) Act 2001 (NSW), it could not be set aside as being beyond authority. That would be an unjust outcome. While it is one possible interpretation on the literal meaning of the words of the section, a consideration of the solemnity of the power to imprison leads inexorably, in my view, to the conclusion that s 5(4) is directed only to the consequences of a failure to comply with the affirmative obligation in s 5(2) to state reasons and does not extend to a breach of the prohibition in s 5(1). Section 5(1) gives effect to a policy that imprisonment is a punishment of last resort. In my view, that policy must be seen as a constraint on the institutional authority of the court which informs the proper construction of the statute.

My conclusion on that issue is fortified by my consideration of the other provisions of the Crimes (Sentencing Procedure) Act and other statutes referred to by Bell P as examples of the same drafting technique of stating that a sentence is not invalidated by a failure to comply with the provision. Most address the consequences of failure to comply with an obligation to give reasons or to explain the effect of an order to the offender. Some address the consequences of failure to comply with a requirement concerning the structure of or process for imposing a sentence. None addresses the decision to imprison.

Against those considerations, I apprehend the principal reason for my disagreement with the majority stems from a different understanding of the task the judge had to perform in the present case. Section 7 of the Crimes (Sentencing Procedure) Act confers a discrete function that is to be performed only after the sentencing judge has determined to sentence an offender to a term of imprisonment. In determining to impose a sentence of imprisonment, the sentencing judge will have performed the task explained in Markarian v R (2006) 228 CLR 357; [2005] HCA 25 at [51] of identifying all the factors that are relevant to the sentence, discussing their significance and making a value judgment as to what is the appropriate sentence given all the factors of the case. As stated in Markarian, it is only at the end of that process that the judge determines the sentence; and it is only then that s 7 comes into play. Of course, many of the factors relevant to the determination of the sentence will also be relevant to the exercise of the power conferred by s 7 (although s 66(1) re-orders the weight to be given to the various relevant factors). My point is that the task is discrete. Section 7(1) provides:

The power conferred by that section assumes a sentence of imprisonment and presents a binary choice as to the manner in which it is to be served. That choice is governed by Part 5, which includes s 66.

I said at the outset of this judgment that I would explain what I mean by “a case in which the issue of the manner in which a sentence of imprisonment is to be served properly arises”. As others have observed, there are statutory constraints on the circumstances in which an intensive correction order is available. The occasion to comply with s 66(2) would never arise in such a case and accordingly those provisions can be put to one side in the present task. I have approached s 66 on the basis that it addresses the circumstance where the power under s 7 is enlivened.

The history of the introduction of intensive correction orders as a new sentencing option is summarised in my judgment in Wany at [3] to [11]. I will not repeat what I said there but the following discussion assumes familiarity with that part of the judgment. In short, intensive correction orders were introduced in 2010 as an option designed “to reduce an offender’s risk of reoffending through the provision of intensive rehabilitation and supervision in the community”. [54] A report tabled in Parliament in 2013 noted that such orders had some significant advantages over full-time imprisonment but despite those advantages the option of an intensive correction order was not being used to any significant extent. [55] The current provisions were introduced following the tabling of that report; they appear to have been intended to encourage greater use of the non-custodial option for serving a sentence of imprisonment. The reforms reflected the government’s commitment to “a tough and smart criminal justice system that puts community safety first”. [56]

The amendments were effected by the Crimes (Sentencing Procedure) Amendment (Sentencing Options) Act 2017. It is clear in my view that Parliament intended by those reforms to confer on the sentencing courts the responsibility to make community safety the paramount consideration when making the binary determination as to how shorter sentences of imprisonment should be served. Further, it is clear that Parliament contemplated that community safety might in some instances be better served by making an intensive correction order (“to reduce an offender’s risk of reoffending through the provision of intensive rehabilitation and supervision in the community”) [57] rather than by committing an offender to prison (a choice noted in the 2013 Report to be more costly and to have “potential contaminating effects” from exposing offenders to other offenders in the prison environment). [58] The Attorney General said in the second reading speech “community safety is not just about incarceration”. [59] He noted evidence showing that “community supervision and programs are far more effective” at reducing offending. [60] These were powerful reforms clearly aimed at reducing the number of people serving short sentences of imprisonment in prison and instead investing the saved resources in rehabilitation and prevention. The paramountcy of community safety was plainly deliberate and cannot be swept aside.

This was a case in which the application of s 7 properly arose. Indeed, the manner in which the sentence of imprisonment was to be served was the only real issue presented for the judge’s determination, all else being conceded by the applicant. The whole focus of the written and oral submissions put to the judge on behalf of the applicant was to persuade the judge that an intensive correction order was more appropriate in the circumstances of the case. The applicant’s lawyer conceded that the s 5 threshold was met and that the indicative sentences specified by the magistrate for the aggregate sentence imposed were within range. The only issue as to which he sought to sway the judge from the course taken by the magistrate was whether an order should be made that the concededly appropriate sentence of imprisonment should more appropriately be served by intensive correction in the community.

The inquiry mandated by s 66(2) is forward-looking. As I have explained, the power under s 7 to make an intensive correction order only arises after the judge has determined that no other sentence than imprisonment is appropriate and has determined the appropriate term of the sentence. The question then is, which method of serving that sentence will best serve the paramount consideration of community safety? That was the entire focus of the submissions put by the applicant’s lawyer. While s 66(2) was not referred to in terms, the applicant’s lawyer would have taken it to be understood that that was the exercise because it was the law he was inviting the judge to apply. He opened his submissions by addressing the decision in Wany, the headnote from which includes the following:

My point here is not to defend my ill-fated judgment in Wany but only to demonstrate that, in my assessment, by relying on that judgment, the applicant’s lawyer made it perfectly clear that he was seeking to persuade the judge that an intensive correction order was more likely to address the applicant’s risk of reoffending. After addressing Wany and making the point that an intensive correction order was available in this case, the lawyer turned to address the findings of fact contended for by him from the applicant’s evidence (set out in the written submissions) and from the Sentence Assessment Report. He then addressed the facts of the offences and continued:

The judge tested that last submission by reference to the applicant’s criminal record and the lawyer conceded certain matters. He then returned to the issue of the manner in which the sentence of imprisonment should be served in the following submissions:

In my view, those submissions were plainly directed to persuading the judge that an intensive correction order was more likely to address the offender's risk of reoffending than serving the sentence by way of full-time detention.

I appreciate that the judge had to exercise the sentencing discretion afresh but in truth the only real issue was whether to accede to the submission that she should make an intensive correction order. Her Honour was required, for that purpose, to make community safety the paramount consideration and for that purpose to comply with s 66(2).

For the reasons Beech-Jones JA has given, I am satisfied that her Honour failed to comply with that section. The judge explained the consideration she had given to community safety as follows:

The suggestion raised by the Court at the hearing that her Honour may have intended to say “posed” rather than “pose” is inconsistent with the previous sentence, which uses the present tense, “There are a substantial number of firearms.” But even if her Honour turned her mind to the obvious fact that those particular guns no longer posed a risk, there is nothing to suggest she undertook the forward-looking task of considering which method of serving the sentence of imprisonment was most likely to address Ms Stanley’s risk of re-offending: two years of full-time detention (removing her from her children and her employment) or a period of supervision in accordance with the plan set out in the Sentence Assessment Report, which was to include fortnightly Cognitive Behavioural Therapy sessions, referrals to alcohol and other drug services and referral to a suitable mental health provider for assessment.

I am further satisfied that the judge’s failure to undertake that task reveals a fundamental misconception of the function conferred by statute, which was to determine the manner in which a determined sentence of imprisonment was to be served after undertaking a real assessment as to which method of serving the sentence of imprisonment would best address the offender’s risk of reoffending. For those reasons, I would have allowed the application.

It remains to address two further matters arising from the other judgments. First, Leeming JA has ventured the view that there is “good reason to think that the assessment required by s 66(2) was of relatively limited significance” in this case. I will not engage with that view as it is, with respect, not relevant to the determination of the application. However, sitting as it does in what will no doubt become an important judgment, I wish to record that I do not agree with what his Honour has written on that subject.

Secondly, I agree with the observations of Beech-Jones JA as to the two matters addressed at the conclusion of his Honour’s judgment.

BEECH-JONES JA: The background to this application is set out in the other judgments. During the course of oral argument, the issues raised by the application were reduced to whether Williams DCJ failed to make the assessment referred to in s 66(2) of the Crimes (Sentencing Procedure) Act 1999 (the “Sentencing Act”) and whether any such failure amounted to a jurisdictional error. Some of the other judgments invert the order in which those issues should be addressed and raise a further issue that was not agitated on the appeal, namely whether, in considering whether to impose an ICO, the sentencing court is required to make an assessment referred to in s 66(2). With the former, the order in which the questions are addressed does not matter. With the latter, the words mean what they state.

In relation to the first issue, although her Honour clearly embarked on a consideration of whether or not to impose an Intensive Corrective Order (“ICO”), I am satisfied that her Honour did not undertake the assessment referred to in s 66(2). The entire focus of the applicant’s appeal from the Local Court to the District Court was whether or not the sentence imposed by Local Court should be served by way of an ICO. The appeal raised a difficult issue for her Honour in addressing the relative seriousness of the offences to which the applicant pleaded guilty against what was a strong subjective case. It was a quintessential case for a consideration of the re-ordered sentencing objectives effected by s 66 (see R v Pullen [2018] NSWCCA 264 at [86] to [89] per Harrison J, with whom Johnson and Schmidt JJ agreed; “Pullen”; Mandranis v R [2021] NSWCCA 97 at [50] to [51] per Simpson AJA with whom Garling and N Adams JJ agreed in this respect; “Mandranis”).

At the hearing of the application, various attempts were made by the respondent to construe her Honour’s reasons so as to warrant the implication that her Honour made the assessment referred to in s 66(2) even if the outcome was not expressly stated in the reasons. All were to no avail. If the assessment had been undertaken and the outcome was that serving the sentence was not more likely, and indeed was far less likely, to address the applicant’s risk of reoffending, then that would have significantly advanced the applicant’s case for an ICO. (Although other sentencing considerations would have remained to be considered and may still have warranted the imposition of a full-time sentence.) However, the analysis never got to that point. The sentencing judge’s consideration never travelled beyond a brief reference to the contents of the Sentencing Assessment Report and the applicant’s past employment history as being “positive indicat[ors] towards good prospects of rehabilitation”. This was stated in a context when, at the hearing of the appeal before her Honour, the applicant gave sworn evidence as follows:

This explanation for the offences was not addressed in the sentencing judge’s reasons. In the context of a consideration of “community safety” and the necessity to make an assessment of whether full time custody is more likely to address the offender’s risk of reoffending, the failure of the sentencing judge to address whether the applicant was a dedicated gun runner or someone caring for five children who just wanted the guns out of her house was significant. If the sentencing judge had made the assessment required by s 66(2) then it can be expected that the resolution of that issue would have been undertaken and produced something more significant than the tepid assessment that she had some “positive indications towards good prospects of rehabilitation.”

In reaching this conclusion I have not overlooked the observation of Adamson J in Mourtada v R [2021] NSWCCA 211 at [37] that “it does not follow from the requirement that matters be considered, that each must be specifically addressed in the reasons given by the sentencing judge”. In some cases, the absence of express reference to some factor or required assessment in a sentencing judge’s reason will not lead to a conclusion that it was not considered, in other cases it will support that conclusion. In particular, the absence of any specific finding that corresponds with s 66(2) in the sentencing judge’s reasons will not necessarily warrant the conclusion that the assessment was not undertaken and considered in the context of s 66(1). The outcome of the assessment may have been a given in the course of the sentencing proceedings or it may otherwise be apparent that the sentencing judge had made the assessment. However, as I have explained, in the context of this case, the conclusion that the assessment was not undertaken is unavoidable.

In relation to the second issue, like Bell P, I do not accept that the sentencing judge’s failure to make the assessment referred to in s 66(2) of the Sentencing Act is a jurisdictional error. In that respect I agree with Quinn v Commonwealth Director of Public Prosecutions [2021] NSWCA 294 (‘Quinn”). In reaching that conclusion, I note that Wany v Director of Public Prosecutions (NSW) (2020) 103 NSWLR 620; [2020] NSWCA 318 (“Wany”) held to the contrary but, with respect, I am satisfied that that aspect of Wany was “plainly wrong” (Gett v Tabet [2009] NSWCA 76; “Gett”). For my part, the critical aspect of the Sentencing Act that warrants that conclusion is that the outcome of any assessment under s 66(2) is not determinative of whether an ICO should or should not be made. Sub-section 66(2) is prefaced by the words “[w]hen considering community safety”. In that sense the assessment referred to in s 66(2) is a component of the consideration of “community safety” dictated by s 66(1) to be a “paramount consideration”. A statutory requirement to consider a particular matter is usually taken as requiring that it be “give[n] weight … as a fundamental element in making” the relevant determination (The Queen v Hunt; ex parte Sean Investments Pty Ltd (1979) 180 CLR 322; [1979] HCA 32] at 329 per Mason J). However, on any view the Sentencing Act does not make the outcome of the assessment in s 66(2) or the consideration of community safety mandated by s 66(1) determinative of a sentencing judge’s consideration of whether or not make an ICO. By contrast, s 67 precludes the granting of an ICO when sentencing for particular offences.

The fact that a particular statutory provision is in this sense determinative is neither sufficient nor necessary to warrant a conclusion that it imposes a jurisdictional precondition to the exercise or non-exercise of a statutory power conferred on an inferior court. However, in this context the fact that it is not determinative is a very strong indicator that is not jurisdictional. Save for Wany, none of the parties nor my research could locate any relevant authority in which a statutory provision which imposed a requirement on an inferior court to “consider” a particular matter has been held to impose a jurisdictional requirement that that consideration be undertaken (see Quinn at [120] and [128] per Leeming JA).

I otherwise agree with the reasons given by Bell P on this issue.

In light of the issues canvassed by some of the other judgments, it is necessary to note two further matters.

First, the relevant authorities concerning the proper approach to s 66 of the Sentencing Act that are binding in the Court of Criminal Appeal are Pullen and Mandranis. To the extent that R v Fangaloka [2019] NSWCCA 173 especially at [60] and [66] to [68] (and its progeny) might be taken as departing from them then, in contrast to the approach taken in this case to Wany, they did not address Pullen in light of the principle that intermediate appellate courts should only depart from their previous decisions if they are of the view that the decision is “plainly wrong” and there are “compelling reasons” to do so (Gett at [273], [277] to [278], [281], [286] and [301]).

Further, as prior decisions of another intermediate court of appeal dealing with the same statutory provisions, Pullen (and Mandranis) should be followed by this Court unless it is convinced that they are clearly wrong (Gett v Tabet supra). Even if this Court did so conclude, it should still follow Pullen to avoid the circumstance that the District Court is bound by one approach when dealing with matters on indictment and another one when hearing matters on appeal from the Local Court (Gett at [296] to [301]) and to provide similar consistent guidance to the Local Court. In an area of great importance to sentencing courts at first instance there are exceptionally strong reasons for this States’ two intermediate courts of appeal to follow their own and each other’s previous decisions.

Second, while a consideration of the “consequences for the parties of holding void every act done in breach” of a statutory condition is relevant to a determination of whether some contravention of a statute amounts to jurisdictional error (Project Blue Sky Inc v Australian Broadcasting Authority (1988) 194 CLR 355; [1998] HCA 28 at [91]), there are limits to any attempt to invoke the spectre of invalidity as a basis for concluding that an error by an inferior court is not jurisdictional. Such reasoning is not a basis for courts below the High Court in the judicial hierarchy to quarrel with the doctrine of jurisdictional error itself rather than its outcome in a particular context. Thus, an asset preservation order made by the District Court was held null and void in Pelechowski v Registrar, Court of Appeal (NSW) (1999) 198 CLR 435; [1999] HCA 19 (esp at [71]) without any apparent or at least much concern for the practical consequences of that finding for other cases. A consideration of the consequences of invalidity met a similar fate in Minister for Immigration and Multicultural Affairs v Bhardwaj (2002) 209 CLR 597; [2002] HCA 11 (“Bhardwaj”) where a submission that the consequence of finding that “[a] decision that involves jurisdictional error is a decision that lacks legal foundation and is properly regarded, in law, as no decision at all” (at [51]) meant that literally thousands of decisions to grant visas could in effect be nullities with all the inconvenience that potentially entailed only found favour with Kirby J (at [122]). In Kirk v Industrial Court of New South Wales (2010) 239 CLR 531; [2010] HCA 1, the jurisdictional errors on the part of the Industrial Court were found to be a failure to identify at any point in the proceeding a matter which was an essential element of the offence charged and the calling of a defendant as a witness contrary to s 17(2) of the Evidence Act (at [73] to [77]). While the court that committed these errors was a superior court of record such that its decisions are valid unless and until set aside, the absence of any suggestion in the judgment in Kirk that that factor tended in favour of a finding of jurisdictional error suggests that the same errors committed by an inferior court would have suffered the same fate.

In this case the spectre of invalidity is exemplified by the suggestion in oral argument that, if the applicant’s argument is correct and the primary judge had imposed an ICO, then the applicant would still be liable to imprisonment as the ICO would be “no decision at all” and she would still be liable to serve the sentence imposed by the Local Court. This line of reasoning should be treated with circumspection. It overlooks the following statement of Hayne J in Bhardwaj at [151]:

Although these observations concern an administrative act, they are equally applicable to the orders of an inferior court that are affected by jurisdictional error. Hence if the applicant had received an ICO then, absent any challenge of the kind stated, the orders releasing her would be presumed to be valid. Further, even if there were proceedings instigated in which there was a direct or collateral challenge to any orders of the District Court releasing her on an ICO, she could still raise the various discretionary considerations that can warrant relief being refused (Re Refugee Review Tribunal; Ex parte Aala (2000) 204 CLR 82; [2000] HCA 57).

Lastly, all the dire warnings about invalidity need to be considered against the potentially serious consequences of not finding jurisdictional error where a judge misapplies or acts contrary to a provision of the Sentencing Act. Thus, if the concerns about the consequences of invalidity were taken to their logical consequence, it would mean that any sentence imposed by a judge who acted on the basis that, contrary to s 21(2) of the Sentencing Act, they must impose the stated (maximum) penalty for an offence would be impervious to challenge (and not capable of correction under s 43).

I agree with the orders proposed by Bell P.