//
© 2026 Zoe. All rights reserved.
Zoe is a legal information platform. Always consult the Federal Register of Legislation for authoritative text.
[2012] NSWCCA 25
appellant. Leave to appeal granted; sentences imposed by Berman DCJ on 18 February 2011 set aside and replaced with an aggregate sentence of 3 years imprisonment, comprising a non-parole period of 16 months...
The Court of Criminal Appeal allowed Rosenstrauss's appeal against the non-parole period imposed by the District Court. Applying the principle that re-sentencing should reflect the law and sentencing practice at the time of the offending, and having regard to historical remission and parole practices, the Court found the two-thirds non-parole period excessive. Using its statutory power to impose an aggregate sentence, the Court set aside the earlier sentences and imposed a three-year aggregate sentence with a 16-month non-parole period, directing parole to follow accordingly.
AI-generated legal information, not legal advice. Zoe can make mistakes. Check the cited source, and for advice about your situation consult a qualified Australian lawyer.
2,116 words · generated 27/06/2026
Jeffrey Stanley Rosenstrauss pleaded guilty to three counts of assault involving an act of indecency upon a person under 16, alleged to have occurred during a period that concluded around May 1981 when the complainant was 11 years old [2]-[3]. He was sentenced in the District Court by Berman DCJ on 18 February 2011 to an overall sentence of three years imprisonment, with a non-parole period of two years, both to date from the sentencing date [4]. The applicant applied for leave to appeal the sentencing order on the ground that the non-parole period did not reflect the sentencing law and practice that would have been applied at the time of the offending, specifically that historical practice would have produced a non-parole period between one-third and one-half of the head sentence [5].
8 cited instruments linked from this judgment.
The Court of Criminal Appeal (Basten JA, RS Hulme J, and Schmidt J) heard the application. Basten JA delivered the principal reasons. The Court applied established authorities that, on re-sentencing after a sentence is quashed on appeal, the appellant is ordinarily to be re-sentenced according to the law and sentencing practice as it stood at the time the sentence was imposed or the offence committed where that law is more favourable to the offender (adopting the Radenkovic principle, applied in R v MJR) [6]-[7]. The Court analysed the operation of remissions and parole in the 1970s and early 1980s, concluding that a sentencing judge in the relevant period would have fixed a non-parole period lower than that imposed by Berman DCJ, because historical practice took into account ordinary remissions and other likely reductions in time served [11]-[16]. The sentencing judge's fixing of a non-parole period equal to two-thirds of the sentence was held to be inconsistent with that historical practice, and therefore erroneous [16].
Because the overall three-year sentence was not challenged by the Crown, the Court faced the task of varying the non-parole part only. In doing so, it recognised (and exercised) the post-2011 statutory power under the Sentencing Procedure Act 1999 (NSW) to impose an aggregate sentence and a single aggregate non-parole period (s 53A(1) and s 44(2A)), a power that had not been available to the sentencing judge in February 2011 but which the Court could use when re-sentencing on appeal [21]-[23]. The Court set aside the District Court sentences and substituted an aggregate sentence of three years imprisonment with a 16-month non-parole period to date from 18 February 2011, directing release on parole on 17 June 2012 [23]-[25].
RS Hulme J and Schmidt J agreed with the orders and the reasons [26]-[29].
Three principal legal concepts shaped the decision:
The Radenkovic/MJR principle: When an appellate court re-sentences after quashing a sentence, considerations of justice and equity ordinarily require re-sentencing according to the law and sentencing practice as it stood when the original sentence was imposed, especially if that law is more favourable to the offender [6]. The Court applied that general principle as binding and directly relevant to the present case [6]-[7].
Historical remission and parole practice: In the period relevant to these offences, remissions operated under regulations made pursuant to the Prisons Act 1952 and the Prisons Regulation 1968, not merely by executive grace. Ordinary remissions varied depending on whether the offender had previously served three months or more (one-quarter remission) or not (one-third remission). The Parole of Prisoners Act 1966 introduced a parole system which interacted with remissions such that anticipated remissions had to be considered in fixing non-parole periods to avoid the absurdity of a prisoner being released on remission before the expiration of a non-parole period fixed by the court (Maclay; Reg v Furey) [11]-[13]. The Court followed authorities showing that in the early 1980s a non-parole period generally would have been fixed at somewhere between one-third and one-half of the head sentence, and that a sentencing judge would expect the head sentence to be reduced by more than one-third when account was taken of ordinary and potential additional remissions [9], [14]-[15].
Statutory re-sentencing powers: Although the District Court judge could not have fixed an aggregate non-parole period in February 2011, the Court of Criminal Appeal, on re-sentencing, could exercise the then-available statutory power in the Sentencing Procedure Act 1999 to impose an aggregate sentence and set one non-parole period for all offences [21]-[22]. That procedural mechanism enabled the Court to implement an aggregate non-parole period consistent with historical sentencing practice while maintaining the overall sentence length.
Applying these principles, the sentencing judge’s choice of a non-parole period equal to two-thirds of the head sentence was inconsistent with what a sentencing judge would have done in the early 1980s. Because the Crown had not challenged the overall three-year term, the Court confined its intervention to varying the non-parole element, substituting a 16-month non-parole period which fell within the historical range identified (12 to 18 months for a first offender) [16], [23]-[24].
Before:
After:
Paragraph 6 (quoting Radenkovic): "In the context of an appeal against sentence ... re-sentenced according to the law as it stood at the time when he was initially sentenced ...", Translation: When an appellate court has to re-sentence someone after overturning the original sentence, it should normally apply the laws and sentencing customs that were in effect when the person was first sentenced, especially if those older rules are better for the offender [6].
Paragraph 11: Explanation that remissions were operated by regulations under the Prisons Act and that entitlement varied (one-quarter or one-third), Translation: Back in the 1970s and early 1980s, reductions in time served were governed by written prison regulations. Those rules typically meant some automatic reduction (remission) of the sentence, so sentencing judges then expected that prisoners would not serve the whole head sentence [11].
Paragraph 12: Courts should not increase punishment to counteract remissions but must take remissions into account when setting non-parole periods to avoid an offender being released before the non-parole period ends, Translation: Judges could not bump up a sentence just to cancel out remissions. Instead they set the non-parole period lower, anticipating remissions, so parole and remission systems worked sensibly together [12].
Paragraph 16: "The sentencing judge imposed a non-parole period which was two-thirds of the sentence period. Because that would not have accorded with sentencing practice at the time of the offending, his Honour erred.", Translation: The District Court judge set a non-parole period that was too long compared to what judges would have set at the time the crimes happened; that was a legal mistake [16].
Paragraph 22: Explanation of s 53A and s 44(2A) powers, Translation: Although the trial judge could not set a single non-parole period for multiple offences in 2011, the Court of Appeal can now use the modern law to impose one aggregate sentence and one non-parole period for all offences when re-sentencing [22].
The decision is directly on point where:
Practically, this precedent is relevant whenever sentencing involves historical offending and a tension exists between contemporary sentencing mechanics and historical parole/remission regimes.
In this judgment the Court relied on, followed or cited several prior authorities (Radenkovic; R v MJR; AJB; Maclay; Smith; Reg v Furey; Mottram; Baxter). The Court distinguished Mottram where factual differences (concurrency/accumulation considerations) meant that Mottram did not require intervention [18]. The approach here reaffirms the line of authority that appellate re-sentencing should reflect historical legal contexts when those contexts are more favourable to the offender [6]-[9]. RS Hulme J acknowledged in separate concurrence that although he had reservations about altering an unchallenged overall sentence, authorities supported that a three-year overall sentence could be appropriate in comparable historical settings [27]-[28]. Practitioners should therefore expect that courts will closely examine historical remission and parole regimes where sentencing relates to past offending and that appellate courts will not hesitate to vary non-parole periods to align with historical practice, using modern statutory aggregate powers when necessary [21]-[25].
Date to be used for historical sentencing practice: The judgment observes a potential fine point about whether courts must apply practice at the date of offending or at the date of likely sentencing (or the actual sentencing date), particularly given changes between Parole of Prisoners Act 1966 and Probation and Parole Act 1983. Basten JA noted the question as 'nice' but proceeded on the basis of practice at the relevant historical time; that subtlety remains a potential issue in other cases where the timing of statutory/regulatory change is close to the period of offending [13].
Extent to which appellate courts should attempt to replicate executive practices: Howie J's passage in AJB, cited and discussed, warns against appellate courts attempting to precisely replicate the executive operation of remissions; the Court followed that caution. The boundary between applying historical sentencing principles and inappropriately attempting to reconstruct executive remission practice may attract future litigation where the precise operation of remissions materially affects sentencing outcomes [8].
Interaction between concurrent/fixed-term structuring and aggregate non-parole setting: Here, the structure of fixed-term sentences with staggered commencements created practical complexities (see [20]). Future cases may require more detailed guidance on how to reconcile historical parole/remission expectations with modern aggregate sentencing forms, particularly where varying commencement dates and partial concurrency produce different temporal overlaps.
Limits of appellate substitution: The Court exercised modern statutory powers (s 53A/s 44(2A)) to effect a historical-consistent non-parole period despite those powers not being available to the sentencing judge. That raises the question of how far appellate courts can or should use post-dating statutory tools when re-sentencing to effect historical policy outcomes. The Court authorised such use here, but varying factual contexts may produce closer scrutiny of the appropriateness of invoking contemporary sentencing mechanics when the original sentencing court could not have used them [21]-[23].
Practitioners re-sentencing historical offending should analyse carefully (a) the statutory and regulatory regime in force at the time of offending and at the time of sentencing, (b) how remissions and parole would have affected the non-parole proportion, (c) whether the Crown challenges the overall sentence, and (d) whether the Court of Criminal Appeal might sensibly use modern aggregate sentencing powers to give lawful effect to the historically appropriate non-parole period.
D L Carroll - Applicant F A Veltro - Respondent Solicitors:
Legal Aid NSW- Applicant S Kavanagh, Solicitor for Public Prosecutions - Respondent File Number(s): CCA 2010/95566 Decision under appeal Jurisdiction: 9101 Citation: R v Rosenstrauss [2011] NSWDC 46 Date of Decision: 2011-02-18 00:00:00 Before: Berman DCJ File Number(s): DC 2010/95566
Judgment 1BASTEN JA : The issue raised by this application for leave to appeal falls within a small compass. 2On 18 February 2011 the applicant was sentenced by Berman DCJ in the District Court for three sexual assaults upon a young girl, committed some 30 years earlier. The precise dates of the offences were not clearly established by the evidence because the complainant was 10 years or younger at the time of the first offence. However, the end of the period of offending appears to have been May 1981, when the complainant was 11 years of age. 3The complainant did not go to the police until 2009. The charges on the indictment, however, alleged offences against the law as it stood up to May 1981, under s 76 of the Crimes Act 1900 (NSW), since repealed. Each charge was of assault involving an act of indecency upon a person under 16 years of age. The maximum penalty for the offence at that time was six years imprisonment. 4It will be necessary to address the specific sentences in due course. Relevantly in respect of the appeal, the effect of the sentencing was to impose an overall period of three years imprisonment, with a non-parole period of two years. Those periods were to date from 18 February 2011, being the date of sentencing. 5The ground of the proposed appeal is that although the sentencing judge sought to apply sentencing practices as they operated at the time of the offences, the non-parole period failed in that respect as at that time a relevant non-parole period would have been between one-third and one-half of the overall sentence period. 6The correct approach, in terms of principle, was established by this Court in R v MJR [2002] NSWCCA 129; 54 NSWLR 368. The Court adopted the principle identified by Mason CJ and McHugh J in Radenkovic v The Queen [1990] HCA 54; 170 CLR 623 at 632: "In the context of an appeal against sentence, when a court of criminal appeal is called upon to resentence because it has quashed the sentence initially imposed, considerations of justice and equity ordinarily require that the convicted person be re-sentenced according to the law as it stood at the time when he was initially sentenced, particularly when that law was more favourable to him than the law as it existed at the hearing of the appeal. The convicted person had an entitlement when he was sentenced by the sentencing judge to a sentence imposed in conformity with the requirements of the law as it then stood." 7Section 19 of the Crimes (Sentencing Procedure) Act 1999 (NSW) ("the Sentencing Procedure Act ") limits the effect of any increase in penalty to offences committed after its commencement. Consistently with that approach, the Court in MJR held that it should take into account sentencing practice as it was at the date of commission of the offence, when sentencing law and practice has changed adversely to an offender: at [27] and [31]. 8In AJB v The Queen [2007] NSWCCA 51; 169 A Crim R 32, Howie J (Adams and Price JJ agreeing) addressed a submission that it was necessary to take into account the fact that, in 1982, being the end of the period of offending in that case, remissions were available to reduce sentences imposed by the courts. Howie J stated at [31]: "There are many cases that indicate how the remission system operated and it is fair to say that, had the applicant been sentenced in 1982, he would, in all probability if not certainty, have received the benefit of a reduction of one third of his sentence. At this time remissions only applied to the head sentence and not to the non-parole period; see generally R v Maclay (1990) 19 NSWLR 112; 46 A Crim R 340 as to the history and application of remissions. But I do not believe that it is appropriate for this Court to try to replicate, not only the sentencing practices of 1982, but also executive practices of that time in respect to the treatment of prisoners." 9Howie J noted that in 1982 "a non-parole period was fixed at somewhere between a third and a half of the term of the sentence": at [39]; see also Dousha v R [2008] NSWCCA 263 at [35] (Fullerton J, Bell JA and Latham J agreeing); Mottram v Regina [2009] NSWCCA 210 at [41] (Hoeben J, Macfarlan JA and Grove J agreeing) and PWB v R [2011] NSWCCA 84 at [64] (RS Hulme J). 10The remission system was abolished with reforms described as reflecting "truth-in-sentencing", effected by the Sentencing Act 1989 (NSW), which commenced on 25 September 1989. 11The remission system operated not pursuant to the general law, or executive act (although there were areas of discretion) but pursuant to regulations made under the Prisons Act 1952 (NSW) and, at the relevant time, the Prisons Regulation 1968 (NSW). The entitlement to "ordinary" remissions varied depending upon whether the offender had previously served a period of imprisonment of three months or more. If so, he or she was entitled to remission of one-quarter of the sentence period; if not, the entitlement to remission was one-third of the sentence period. (The relevant provisions are conveniently found in Smith v Corrective Services Commission (NSW) [1980] 2 NSWLR 171 at 183-185, in the judgment of Hope JA; see also Smith v Corrective Services Commission (NSW) [1980] HCA 49; 147 CLR 134 and Green v Corrective Services Commission (NSW) [1982] 1 NSWLR 327.) 12As explained by Gleeson CJ, Hunt and Loveday JJ in Maclay , it was established principle that the sentencing judge should not increase the sentence to counteract the effect of anticipated remissions on the period to be served. However, that did not mean that the court ignored the anticipated remissions. With the introduction of the parole system pursuant to the Parole of Prisoners Act 1966 (NSW), it became necessary for the sentencing judge to take account of the likely diminution on account of remissions in order to avoid "the absurd result that a prisoner would be, as a result of the remissions system, released before the expiration of the non-parole period specified by the sentencing judge, with the result that the non-parole period would be rendered nugatory": Maclay at 117G; Reg v Furey [1971] 2 NSWLR 262 at 267. 13There might be a nice question as to whether it would now be necessary to consider the sentencing practices at the date of the offending, or as at the date of likely sentencing. With the commencement of the Probation and Parole Act 1983 (NSW), the absurdity which had previously existed was removed by providing for remissions to apply to the non-parole period. The result was, again in the words of the Court in Maclay, that "all marched precisely in step in a proportionate sense": at 118G. However, the courts held that "neither in relation to head sentences nor in relation to non-parole periods were sentencing judges to increase terms to negative the effect of remissions that would be applied to them": Maclay at 120D. 14Applying the law, rather than executive practice, a sentencing court, whether in 1981 or 1984, would have fixed a non-parole period on the assumption that the overall "sentence period" would be reduced not only by one-third, on account of ordinary remissions, but also for "earned" remissions and, in all probability, other remissions which might accrue from time to time, such as "special strike remissions", designed to recognise the increased harshness of imprisonment during periods of industrial action by prison officers, which were not uncommon. 15By way of example, the offender in Smith's case, who was serving an overall period of 12 years on account of five cases of rape, and was not a first offender, received a non-parole period of seven years, being 58% of the sentence period. Because, in the present case, a sentencing judge would have expected the applicant's sentence period to be reduced by more than one-third, it is likely that the non-parole period would have been less than 50% of the sentence period. 16The sentencing judge imposed a non-parole period which was two-thirds of the sentence period. Because that would not have accorded with sentencing practice at the time of the offending, his Honour erred. In accordance with the practice identified by Howie J in AJB , the non-parole period would have been between 12 and 18 months in the case of a first offender. (The fact that, if sentenced after the commencement of the Probation and Parole Act 1983 , that period would have been reduced on account of remissions, would have properly been ignored.) 17The remaining question is whether this Court should intervene on the basis that some other sentence is "warranted in law and should have been passed", for the purposes of s 6(3) of the Criminal Appeal Act 1912 (NSW). 18In the course of argument, the Court's attention was directed to Mottram , where a similar error to that identified in the present case had been made in fixing the non-parole period, but the Court did not intervene. In that case the applicant had obtained the benefit of total concurrency in respect of six sentences, where at least in relation to two particular sentences involving a different victim at a later point in time, some degree of accumulation was required: at [47]. Hoeben J concluded that a "proper application of the principles of proportionality and totality to the offending conduct in this case, even allowing for the lower maximum sentences applying during the 1970s and 1980s, would have inevitably resulted in sentences greater than those passed by his Honour": at [49]. However, that passage demonstrates why, on its facts, Mottram was distinguishable from the present case. Not only was there no challenge by the Director to the overall sentence period in this case, but there was also no contention that it should have been treated as manifestly inadequate for the purpose of declining to intervene pursuant to s 6(3). The respondent's contention that no other sentence was warranted in law had regard only to the non-parole period. 19Once the overall sentence period is accepted, and the relevant proportion for the non-parole period is identified in accordance with sentencing practice at the time of the offending, it is necessary to intervene to vary the non-parole period. The phrase "warranted in law" requires that "the entire body of legal rules that inform the exercise of a sentencing discretion, ie both statutory requirements and sentencing principles developed at common law" be applied: Baxter v R [2007] NSWCCA 237; 173 A Crim R 284 at [18] (Spigelman CJ, Latham J agreeing). Critically, in determining the non-parole period, that must require application of the principles applicable at the time of offending, as discussed above. It follows that the non-parole period fixed by the sentencing judge was not that which was warranted by law. 20The respondent raised a separate contention, based on the structure of the sentences imposed by the District Court. Thus, on each of the second and third counts, the applicant was sentenced to a fixed term of one year, with the first term (count 2) to commence on the date of sentencing, namely 18 February 2011, and the second (count 3) to commence six months later, on 18 August 2011. Thus, the second fixed term would expire on 17 August 2012. As a result, 18 months of the period of mandatory custody was to be attributed to two fixed term sentences. The third sentence (count 1) involved a non-parole period of one year, to commence on 18 February 2012 and expire on 17 February 2013. In the result, the actual non-parole period was 50% of the sentence imposed on count 1. However, it constituted only six months of the two year period of mandatory custody. The submissions of the respondent must be accepted in part: if the period of mandatory custody is to be reduced to less than 50% of the total sentence period, it will be necessary to interfere with at least the fixed term imposed in respect of count 3. 21The respondent's submission otherwise reveals a further error on the part of the sentencing judge. Pursuant to the Parole of Prisoners Act 1966 (NSW), as in force in 1982, a non-parole period was to be fixed in respect of the period of imprisonment being served by the offender at the date of sentencing, which might include an accumulation of earlier sentences and new sentences. The period was not fixed in respect of a specific sentence. Accordingly, the correct course in the present case, if dealt with in 1982, would have been to fix the various sentences of imprisonment imposed for the three counts and then specify a non-parole period which would commence on the date of imprisonment and continue regardless of the expiration of a particular fixed term sentence. The practical difficulty which is then apparent is that the sentence being imposed in 2011 is imposed in accordance with the powers conferred by the Sentencing Procedure Act . At the time of sentencing, the Act did not permit an aggregate non-parole period for all offences. 22Since 14 March 2011, a court sentencing an offender for more than one offence may impose "an aggregate sentence of imprisonment": Sentencing Procedure Act , s 53A(1). Although that power was not available to the sentencing judge, it is available to this Court for the purpose of re-sentencing the applicant. In exercising that power, the Court may set one non-parole period for all the offences to which the sentence relates after setting the term of the sentence: s 44(2A). The requirement not to set a balance of term exceeding one-third of the non-parole period, absent a finding of special circumstances, applies in relation to the aggregate sentence: s 44(2B). It is not necessary to indicate what the non-parole period would have been for each separate offence had separate sentences been imposed (s 44(2C)) but the Court is required to indicate to the offender, and make a record of, the sentence that would have been imposed for each offence had separate sentences been imposed: s 53A(2). 23The appropriate course is for the Court to set aside each of the sentences imposed by the sentencing judge and, in place of those sentences, impose an aggregate sentence of three years imprisonment with a non-parole period of 16 months. 24In taking that step, the finding of special circumstances made by the sentencing judge should be accepted as appropriate, for the reasons he gave. Further, the sentences which he imposed in relation to counts 2 and 3, being fixed terms of one year each to commence on 18 February 2011 and 18 August 2011 respectively are the sentences which would have been imposed if an aggregate sentence had not been imposed. In relation to count 1, the sentence of imprisonment which would otherwise have been imposed is a sentence of two years imprisonment to date from 18 February 2012, without setting a non-parole period in relation to that sentence. The appropriate non-parole period in relation to the aggregate sentence is 16 months to date from 18 February 2011. 25The Court should make the following orders: (1) Grant leave to appeal and set aside the sentences imposed by Berman DCJ on 18 February 2011. (2) Sentence the applicant in relation to the offences to which he pleaded guilty to an aggregate sentence of imprisonment for 3 years, comprising - (a) a non-parole period of 16 months to date from 18 February 2011 and to expire on 17 June 2012, and (b) a balance of term of 20 months, commencing on 18 June 2012 and expiring on 17 February 2014. (3) The Court directs the release of the applicant on parole on 17 June 2012. 26RS HULME J : I agree with the orders proposed by Basten JA and with his Honour's reasons. 27I acknowledge to having had some reservations about the result, proceeding as it does from an assumption that the overall sentence was appropriate. That sentence was, as Basten JA has observed, not challenged by the Crown. 28However, it must be accepted that the penalties provided by Parliament at the time of the Applicant's offending and which the Courts were obliged to follow were very much lighter than now and there are cases which support the view that a total of three years was within the proper exercise of Berman DCJ's sentencing discretion - see AJB v R [2007] NSWCCA 51; (2007) 169 A Crim R 32; Bradbery v R [2008] NSWCCA 93; (2008) 184 A Crim R 483; GRD v R [2009] NSWCCA 149. Arguably, some of the cases on s 61E referred to in the table attached to my reasons in PWB v R [2011] NSWCCA 84 lead to the same conclusion. 29SCHMIDT J : I agree with both Basten JA and RS Hulme J.
DISCLAIMER - Every effort has been made to comply with suppression orders or statutory provisions prohibiting publication that may apply to this judgment or decision. The onus remains on any person using material in the judgment or decision to ensure that the intended use of that material does not breach any such order or provision. Further enquiries may be directed to the Registry of the Court or Tribunal in which it was generated. Decision last updated: 05 March 2012
Rosenstrauss
R
Leave to appeal granted; sentences imposed by Berman DCJ on 18 February 2011 set aside and replaced with an aggregate sentence of 3 years imprisonment, comprising a non-parole period of 16 months to date from 18 February 2011 (to expire 17 June 2012) and a balance of term of 20 months (commencing 18 June 2012 to expire 17 February 2014); the Court directed release on parole on 17 June 2012 (see [25]).