Markarian v R [2005] HCA 25; (2006) 228 CLR 357; (2005) 215 ALR 213; (2005) 79 ALJR 1048 (18 May 2005)

Last Updated: 18 May 2005

HIGH COURT OF AUSTRALIA

ANTHONY VASKEN MARKARIAN APPELLANT

AND

THE QUEEN RESPONDENT

1. Appeal allowed.

2. Set aside the sentence and orders of the Court of Criminal Appeal of the Supreme Court of New South Wales.

3. Remit the matter to the Court of Criminal Appeal of the Supreme Court of New South Wales to dispose of the appeal in accordance with the reasons for judgment of this Court.

On appeal from the Supreme Court of New South Wales

Representation:

A C Haesler and R W Burgess for the appellant (instructed by Legal Aid Commission of New South Wales)

R D Cogswell SC with G E Smith and J A Quilter for the respondent (instructed by Solicitor for Public Prosecutions (New South Wales))

Notice: This copy of the Court's Reasons for Judgment is subject to formal revision prior to publication in the Commonwealth Law Reports.

CATCHWORDS

Markarian v The Queen

Criminal law - Sentence - Principles - Drug offence - Appellant acted as driver for heroin dealer - Appellant pleaded guilty and asked that four other offences be taken into account by sentencing judge - Whether Court of Criminal Appeal adopted impermissible approach to sentencing by means of staged approach - Whether staged approach to be preferred to instinctive synthesis of sentencer - Relevance of maximum available sentence - Relevance of quantity of drug.

Criminal law - Appeal - Prosecution appeal against sentence - Court of Criminal Appeal increased sentence from 2 years and 6 months to 8 years - Whether Court of Criminal Appeal was wrong to find that the original sentence was manifestly inadequate - Whether re-sentencing discretion miscarried.

Criminal law - Sentence - Re-sentencing - Further offences - Additional discrete sentence added to head sentence for further offences disclosed by offender - Whether such approach a breach of totality principle.

Criminal law - Sentence - Principles - Failure by trial judge and Court of Criminal Appeal to consider an obligatory requirement of sentencing statute in determining appellant's sentence - Whether sentencing discretion of trial judge and Court of Criminal Appeal miscarried because of such omission.

Words and phrases - "staged approach", "two-stage approach", "instinctive synthesis".

Crimes (Sentencing Procedure) Act 1999 (NSW), Div 3 Pt 3, s 21A, s 31, s 32, s 34(1), s 101A.

Criminal Appeal Act 1912 (NSW), s 5D.

Drug Misuse and Trafficking Act 1985 (NSW), s 33(2).

1. GLEESON CJ, GUMMOW, HAYNE AND CALLINAN JJ. The question in this case is whether the Court of Criminal Appeal of New South Wales failed to apply or misapplied orthodox sentencing principles in upholding an appeal against sentence by the Crown.

   Facts

2. At his arraignment on 3 May 2002, the appellant pleaded guilty to a charge that between 18 April and 10 October 2000 he did knowingly take part in the supply of a prohibited drug, namely heroin, in an amount not less than the commercial quantity for that drug - 415 grams pursuant to s 33(2) of the Drug Misuse and Trafficking Act 1985 (NSW) ("the Act") ("the principal offence"). He asked that in sentencing him for the principal offence four other matters ("the further offences") be taken into account by the sentencing judge. The way in which the further offences should be dealt with is governed by a special statutory regime to which some detailed reference is necessary and will be made later.

3. On 15 April 2002, before the arraignment, s 21A of the Crimes (Sentencing Procedure) Act 1999 (NSW) ("the Sentencing Act") had commenced[1]. Section 21A, headed "General sentencing principles", was not referred to by the sentencing judge or by the Court of Criminal Appeal. This may have been on the assumption of counsel that insofar as it applied to the present case it did not alter the general law principles which otherwise applied. No contrary submission, requiring further attention to s 21A, was made in this Court and, accordingly, its terms need not be considered.

4. The most serious of the further offences were the supply of in excess of 5 grams of heroin between 25 September and 1 October 2001, and of 232.5 grams of cannabis leaf, another prohibited drug, on 29 September 2000.

5. On 18 July 2002 Hosking DCJ sentenced the appellant to a term of imprisonment of 2 years and 6 months from 18 July 2002 with a non-parole period of 15 months. His Honour was of the opinion that the appellant's plea of guilty had utilitarian value. He accordingly discounted the sentence by 25%. The appellant was therefore eligible for release on parole on 17 October 2003. His head sentence was to expire on 17 January 2005. In August 2002 however the respondent appealed.

6. The facts constituting the principal offence consisted of the appellant's acting over a period of 5 months as a driver for Vincent Caccamo, a dealer in heroin. The appellant, who was himself a heroin addict, was paid in heroin for his services. The material before the sentencing court emphasized the different degree of criminality of the appellant from Caccamo's. Caccamo had previously been sentenced to 8 years imprisonment with a non-parole period of 5 years for a number of offences of supply in the course of an illicit business of handling and selling drugs. The relative brevity in all of the circumstances of his sentence is explained by the significant value that the judge who sentenced him attached to his cooperation with the police. Another of Caccamo's drivers, Chung, was sentenced to periodic detention of 3 years with a 2 years period of non-parole. Chung did not have a criminal record. He had fewer other matters to be taken into account, and he had driven less frequently for Caccamo than had the appellant.

7. The appellant gave evidence at the sentence hearing. This, in summary, was that he was born in December 1963 and started to use heroin soon after his mother's death in August 1996. Caccamo became his source for the drug. In April 1998, he was sent to prison. By the time of his release in October 1999 he had taken himself off both heroin and methadone. He however resumed contact with Caccamo in about July 2000. He regarded himself as indebted to Caccamo for the latter's kindness to his father when he was in prison. At this point the appellant resumed drug taking. Caccamo, who did not have a valid driver's licence, used the appellant as his driver in return for drugs. Before he was charged the appellant had dissipated, largely on illegal drugs, an inheritance from his father of $200,000. He claimed that his own criminal activities had been done out of desperation and in despair at the loss of his parents.

8. The appellant has a criminal history. He was placed on recognisance of 3 years for cultivating a prohibited plant and fined for possessing a prohibited imported drug in 1991. In May 1998 he was sentenced to imprisonment for supplying a prohibited drug. For that offence he spent 18 months in prison and an additional 18 months on parole. He was on parole at the time of the commission of the principal offence and one of the further offences.

9. The sentencing judge had before him an optimistic pre-sentence report indicating that the appellant had been in regular employment until about 1990. He had apparently made genuine progress towards drug rehabilitation by the time of sentence.

10. The appeal to the Court of Criminal Appeal was upheld[2] (Hulme J with whom Heydon JA and Carruthers AJ agreed). A sentence of 8 years imprisonment with a non-parole period of 4 years and 6 months was imposed in lieu of the earlier sentence of 2 years and 6 months with a 15 month non-parole period. The appellant is now eligible for release on parole on 18 January 2007. His sentence will expire on 17 July 2010.

11. In his reasons for judgment Hulme J referred to relevant penalties imposed under the Act[3]:

    "The Drug Misuse and Trafficking Act provides for a variety of maximum periods of imprisonment, depending on the quantity and type of drug involved. In the case of the supply, or knowingly take part in the supply, of heroin, the periods are:

    (i) where the quantity is not more than 1 g (a 'small quantity'), and the matter is dealt with summarily, two years imprisonment (s 30);

    (ii) where the quantity is not more than 5 g (an indictable quantity), and the matter is dealt with summarily, two years imprisonment (s 31);

    (iii) where the quantity is less than 250 g and the matter is dealt with on indictment, 15 years imprisonment (s 32);

    (iv) where the quantity is not less than 250 g but not as much as 1,000 g (a 'commercial quantity'), 20 years imprisonment (s 33(2)); and

    (v) where the quantity is not less than 1,000 g (a 'large commercial quantity'), life imprisonment (s 33(3))."

    As to these his Honour observed[4]:

    "Although this summary makes it clear that the maximum sentences prescribed are not proportional to quantities, it is clear that, all other things being equal, Parliament intended that the greater the quantity, the higher the sentence should be. Of course, that is not to say that all other matters relevant to sentence should not also have their proper weight[5]."

    His Honour then referred to other judgments[6] of the Court of Criminal Appeal of New South Wales in which statements drawing attention to the need in sentencing to deter criminal conduct, and to protect the public, without losing sight of tailoring the sentence to the particular circumstances of the offence charged, and to the ensuring of "reasonable proportionality" in that regard, have been made.

12. Hulme J was of the opinion that not one of the principles reflected in the statements to which we have referred was applied by the sentencing judge. He was influenced by his own experience as a trial judge. He said[7]:

    "Much, if not most of the work of the courts is taken up with the consequences of the ravages drugs, particularly heroin, inflict on those who take it and, by them, on society. The survey of imprisoned burglars reported in 'The Stolen Goods Market in New South Wales' conducted by the New South Wales Bureau of Crime Statistics and Research indicated a median expenditure by heroin users of $1,500 per week and the need to steal goods worth a number of times this amount to feed their habit. On average each such offender is thus costing the community through property losses and the like $200,000 per year. And that says nothing about the violence other offenders resort to, or the waste of life and degradation heroin inflicts on the lives of the tens of thousands of persons it comes to dominate. To punish those who help to perpetuate such consequences by sentences such as was imposed in this case is to fail to adhere to the dictates of Parliament, to fail to adhere to basic principles of sentencing, to fail to provide much disincentive to others tempted to offend in the same way, and to fail the community's entitlement to retribution or, as I think is encompassed within that expression, to feel justice has been done."

    His Honour then turned his mind to the particular circumstances of the offence[8]:

    "'The degree by which, having regard to the maximum penalties provided by the Act in question, the [appellant's] conduct ... offend[ed] against the legislative objective of suppressing the illicit traffic in the prohibited drug' was substantial. Albeit it was a long way short of the 999.9 g maximum for a commercial quantity, the 415 g the distribution of which he assisted well exceeded the 250 g upper limit for an indictable quantity for the supply of which Parliament had seen fit to prescribe a maximum penalty of 15 years. The [appellant's] activities extended over a period of almost six months. They amounted to conscious deliberate criminality, day after day, for reward, even if that reward was in the form of drugs. At the time he was on parole - a seriously aggravating feature - and had previously been convicted of supplying prohibited drugs and imprisoned. By his repeated offending the [appellant] 'manifested ... a continuing attitude of disobedience of the law'[9].

    ...

    There were also the offences on the Form 1. The second, involving cannabis, carried a penalty of two or 10 years also depending on whether it was prosecuted summarily or on indictment. In that this offence may have been part of the [appellant's] active assistance to Mr Caccamo in the latter's drug dealing activities, it is proper to regard it as part of the same criminal activity. However, in that a different drug was involved, the criminal activity covered a broader spectrum and merited an increase in punishment.

    It seems likely that the third of the offences in the Form 1 - which being of possession rather than supply, carried a maximum period of imprisonment of two years - was associated with either the operation which was the subject of the first Form 1 offence or the [appellant's] own heroin addiction. It seems very likely that the fourth of these offences, which, under s 527C of the Crimes Act 1900 (NSW) carried a maximum period of imprisonment of six months - was also an incident of the [appellant's] own commercial dealing. It does not disadvantage the [appellant] to so regard both of these offences."

13. Hulme J did not overlook the subjective matters to which regard should be had, the first of which was the appellant's pleas to the principal offence and the further offences. He pointed out however that these were only entered after the committal, and that the evidence against the appellant which included the results of surveillance and taped records of conversations, was strong. His Honour then had regard to the appellant's apparent contrition, his addiction and his attempts to rehabilitate himself. All of these had to be weighed, his Honour said, with the compelling counter consideration, of two previous offences of supplying heroin. And, despite his reservations about the correctness of an earlier line of authority in the New South Wales Court of Criminal Appeal[10] holding that there should be a significant discount in any additional penalty on account of further offences which were admitted, Hulme J accepted that he must give effect to those authorities in this case. His Honour also had regard to the need to apply the totality principle.

14. Later in his judgment his Honour again referred to his personal experience as a trial judge[11]:

    "There is some weight of authority in favour of sentences being determined by instinctive synthesis[12]. However, as one who has had to carry out the sentencing task both in this Court and at first instance, and to examine innumerable sentences imposed by others, my experience is that there are far more advantages in reasoning to a conclusion. I confess that in a significant number of the cases which come to this Court, the instinctive synthesis approach adopted in the cases under appeal have made me wonder whether figures have not just been plucked out of the air. Indeed that is what seems to have occurred in this case. His Honour, having referred to the objective and subjective features, including the [appellant's] addiction and efforts towards rehabilitation, having expressed the opinion that the [appellant] was entitled to a discount of 25% for his plea of guilty, and that he proposed to extend some leniency because Chung was treated very leniently, said simply 'In my view, the starting point for this sentence would have been a sentence in the order of three and a half years but with the offender's plea of guilty, that translates to a sentence of 30 months, namely, two and a half years'.

    No reasons were advanced in support of the three and a half years figure or to explain why it was not five, or seven or 10 years. I acknowledge that, in many sentencing exercises, there will be an element of subjective choice or value judgment which it may be impossible to avoid but it seems to me far preferable that reasoning be apparent in respect of the more significant features than occurred in relation to what was his Honour's fundamental starting point.

    Neither does it seem very satisfactory for me, sitting on appeal, simply to say 'His Honour's instinctive synthesis was manifestly wrong. My instinctive synthesis leads to the view his starting point should have been five (or seven or 10) years'."

15. Intervention by the Court of Criminal Appeal was warranted, his Honour said[13], for these reasons:

    "Had the [appellant's] offence and circumstances fallen within the category of a worst case falling within the statutory provisions, the sentence should have been not less than the 15 years maximum for the offence of supplying an indictable quantity. I appreciate that the charge specified a commercial quantity, that the maximum period of imprisonment prescribed for that offence is 20 years and that the quantity involved in this offence was only a little more than 40% of the maximum commercial quantity. However, Parliament cannot have intended that, other things being equal, the penalty for supplying more than 250 g should be less than for supplying that quantity. The absence of proportionality in the maximum sentences prescribed is perhaps partly explicable upon the basis that the severity of imprisonment is not simply proportional to its length. Having regard to the sorts of terms under consideration for drug dealing a sentence of one of the longer periods is liable to have an impact on an offender's life in terms of wife, children, job prospects and the like far greater than a sentence, say half as long.

    But be that as it may, in face of the totality of the statutory provisions and the principles for which I have cited Veen v The Queen (No 2)[14] and R v Peel[15], it seems to me that the maximum prescribed for the supply of 250 g is not too high a starting point. In R v Perrier (No 2)[16], the Victorian Court of Criminal Appeal thought the sentences prescribed for lesser quantities relevant to the sentence appropriate for higher quantities."

16. After referring again to the subjective factors which he had earlier noted in his judgment his Honour said this[17]:

    "So far as the [appellant's] role is concerned, he was of course not the principal and the charge was not to supply but only of being knowingly involved in supply. While at times he seems to have been no more than a chauffeur, on other occasions his role was substantially more significant. In light of the matters referred to in this paragraph, I would reduce my 15 year starting point by about one-third. A number of factors lead me to the view that the reduction on this account should not be greater. These include the sorts of considerations spoken of in Le Cerf[18]. They include also my view that the severity of sentences is not simply proportionate to length. They include the nature of the [appellant's] activities and the fact that they extended over a much longer period than that during which a courier is normally involved. The conclusion derives some support also from the relativity between the maximum sentences available for importing heroin and the pattern of sentences imposed on couriers involved in the importation of quantities in the top part of trafficable quantities of that drug. Based on the decisions to which I referred, I concluded in R v Spiteri[19] that nine years out of a maximum of 16.5 or 17 years seemed to be the pattern. Of course those figures show a greater difference than one-third and if the comparison is to be made between the circumstances here and those in the cases to which I referred it is necessary to recognise the differences. I need not detail these but they include that most of those cases included a plea of guilty and I am not at this stage taking any account of such a plea of guilty. One must recognise also the difference in nature and extent of the [appellant's] activities compared with those of persons regarded as couriers."

17. Despite his reservations about the appropriateness of a discount of as much as 25%, Hulme J was disposed to allow such a discount to any sentence that he concluded should be imposed, in part at least because the appeal was a Crown appeal. That discount should, he said, be regarded however as the sum of all discounts which might otherwise be made in consequence of the plea and the appellant's admission of guilt to the further offences.

18. In dealing with those offences his Honour said[20]:

    "Operating in the other direction are the offences on the Form 1. I have said sufficient to indicate my view about them save and except that principles of totality have also to be taken into account. On account of the matters on the Form 1, particularly the first and second of these, I would increase the sentence otherwise appropriate by between 18 months and two years."

19. Because of the great disparity in criminality between Chung and the appellant, his Honour did not think that any question of parity of penalties arose. His Honour was however conscious that the sentence that he would substitute was a high one[21]:

    "The second of these further topics to which I should refer arises from the statistics kept by the Judicial Commission. By comparison with those statistics, the sentence I propose is a high one. Those statistics show that, in the period from July 1995 to December 2001, there were 22 offenders sentenced in respect of the offence of being knowingly concerned in the supply of (not less than) a commercial quantity of heroin. Twenty one were sent to prison. The longest full term of imprisonment was eight years, imposed on two offenders, and only seven offenders received full terms of six years or more. The statistics indicate that all the persons referred to in the last sentence pleaded guilty and, except for one of the seven, had matters on a Form 1.

    This comparison raises the question whether the sentence I propose is the lowest which should properly be imposed for the [appellant's] offending but, having reflected on the question, I am satisfied that nothing less will properly reflect the considerations to which I have referred."

    The appeal to this Court

20. In this Court the appellant argues that the Court of Criminal Appeal erred by adopting a staged approach to the calculation of the sentence, in taking a maximum penalty as a starting point for that calculation, and, by, in reality impermissibly imposing a separate penalty for the other offences. Further, the appellant contends that the sentence was, in any event, so plainly unjust that an error in the sentencing discretion was to be inferred.

21. The respondent argued that if at all, it was only in form and not in substance, that Hulme J embarked upon an approach of a two-tiered or sequential kind. This, the respondent accepted, appeared from the order and way in which his Honour dealt with the relevant matters: the statutory provisions and the purposes of them; the worst category of cases; deterrence and public security; proportionality; subjective matters; the circumstances of the offences; criminal history; double jeopardy on a Crown appeal; and the need or otherwise for parity and comparable sentences. Nonetheless, the respondent submitted, his Honour, having taken all possible relevant issues and matters into account, had not been shown to have erred.

22. The appellant submitted that Hulme J fell into error in taking as his starting point, the quantity of heroin the subject of the principal offence. He acknowledged that quantity was clearly a relevant, but contended that it was not the determinative factor. In this connexion he cited the joint judgment of Gaudron, Gummow and Hayne JJ in Wong v The Queen[22]:

    "These are reasons enough for concluding that the Court of Criminal Appeal was in error in attributing chief importance to the weight of narcotic in fixing sentences for the offence. The error of the Court is, however, more deep seated than the factual difficulties to which reference has been made. The selection of weight of narcotic as the chief factor to be taken into account in fixing a sentence represents a departure from fundamental principle."

23. Hulme J, the appellant argued, then engaged in an impermissible arithmetical process after rejecting what has been called an approach by way of instinctive synthesis. His Honour proceeded by referring to a maximum penalty of 15 years, reducing that period by a third because the appellant's role was of a lesser kind than that of Caccamo, making a further reduction of 25% on account of the utilitarian value of the plea and contrition, increasing the sentence by 18 months to 2 years because of the further offences, and taking into account various other factors pointing in different directions, the prospects of rehabilitation, deterrence, the security of the community, and the double jeopardy arising by reason of a Crown appeal.

    The decision

24. It is not useful to begin by asking a general question like was a "staged sentencing process" followed. That is not useful because the expression "staged sentencing process" may mean no more than that the reasoning adopted by the sentencer can be seen to have proceeded sequentially. Or it may mean only that some specific numerical or proportional allowance has been made by the sentencer in arriving at an ultimate sentence on some account such as assistance to authorities or a plea of guilty. Neither the conclusion that a sentencer has reasoned sequentially, nor the observation that a sentencer has quantified the allowance made, for example, on account of the offender's plea of guilty, or the offender's assistance to authorities, of itself, reveals error. Indeed provisions like s 21E of the Crimes Act 1914 (Cth) may require the sentencer, in some circumstances, to identify the amount by which a sentence has been reduced on some account.

25. As with other discretionary judgments, the inquiry on an appeal against sentence is identified in the well-known passage in the joint reasons of Dixon, Evatt and McTiernan JJ in House v The King[23], itself an appeal against sentence. Thus is specific error shown? (Has there been some error of principle? Has the sentencer allowed extraneous or irrelevant matters to guide or affect the decision? Have the facts been mistaken? Has the sentencer not taken some material consideration into account?) Or if specific error is not shown, is the result embodied in the order unreasonable or plainly unjust? It is this last kind of error that is usually described, in an offender's appeal, as "manifest excess", or in a prosecution appeal, as "manifest inadequacy".

26. Any consideration of alleged error of principle must now begin in any applicable legislation governing sentencing either generally or in the particular case. In sentencing for a federal offence, it must begin by considering Pt 1B of the Crimes Act. In the present case, it must begin with the provisions of the Sentencing Act.

27. Express legislative provisions apart, neither principle, nor any of the grounds of appellate review, dictates the particular path that a sentencer, passing sentence in a case where the penalty is not fixed by statute, must follow in reasoning to the conclusion that the sentence to be imposed should be fixed as it is. The judgment is a discretionary judgment and, as the bases for appellate review reveal, what is required is that the sentencer must take into account all relevant considerations (and only relevant considerations) in forming the conclusion reached. As has now been pointed out more than once, there is no single correct sentence[24]. And judges at first instance are to be allowed as much flexibility in sentencing as is consonant with consistency of approach and as accords with the statutory regime that applies[25].

28. The proceedings in the Court of Criminal Appeal being a prosecution appeal, brought pursuant to s 5D of the Criminal Appeal Act 1912 (NSW), it was, of course, necessary for the prosecution to show error in the sentence passed below - either specific error or manifest inadequacy. As the whole Court pointed out in Lowndes v The Queen[26], a court of criminal appeal may not substitute its own opinion for that of the sentencing judge merely because the appellate court would have exercised its discretion in a manner different from the manner in which the sentencing judge exercised his or her discretion.

29. In the present case, contrary to what ordinarily would be expected, the Court of Criminal Appeal did not state explicitly that it was of the view that the sentence below was manifestly inadequate. It was nonetheless apparent from the order ultimately made by the Court that it had reached this conclusion. The appellant submitted in this Court that the Court of Criminal Appeal was wrong to find that the sentence originally passed was manifestly inadequate. Because, for the reasons that follow, the re-sentencing discretion of the Court of Criminal Appeal miscarried, it will be necessary for the matter to be remitted to the Court of Criminal Appeal. It is not necessary in those circumstances to decide whether the original sentence was manifestly inadequate. That will be a matter for the Court of Criminal Appeal to consider afresh in the further hearing of the prosecution appeal.

30. Legislatures do not enact maximum available sentences as mere formalities. Judges need sentencing yardsticks. It is well accepted that the maximum sentence available may in some cases be a matter of great relevance. In their book Sentencing, Stockdale and Devlin[27] observe that:

    "A maximum sentence fixed by Parliament may have little relevance in a given case, either because it was fixed at a very high level in the last century ... or because it has more recently been set at a high catch-all level ... At other times the maximum may be highly relevant and sometimes may create real difficulties ...

    A change in a maximum sentence by Parliament will sometimes be helpful [where it is thought that the Parliament regarded the previous penalties as inadequate]."

31. It follows that careful attention to maximum penalties will almost always be required, first because the legislature has legislated for them; secondly, because they invite comparison between the worst possible case and the case before the court at the time; and thirdly, because in that regard they do provide, taken and balanced with all of the other relevant factors, a yardstick. That having been said, in our opinion, it will rarely be, and was not appropriate for Hulme J here to look first to a maximum penalty[28], and to proceed by making a proportional deduction from it. That was to use a prescribed maximum erroneously, as neither a yardstick, nor as a basis for comparison of this case with the worst possible case. That he used the maximum penalty impermissibly appears from his Honour's particular deference to it in this passage[29]:

    "Parliament cannot have intended that, other things being equal, the penalty for supplying more than 250 g should be less than for supplying that quantity."

    The form of the statement is explained by the fact that his Honour did not start with the maximum penalty for an offence involving the quantity in question, but used another maximum penalty as his starting point, that is, the maximum for an offence in the category of seriousness immediately below that of the principal offence.

32. The appellant's submission that the passage just quoted involved too great an emphasis upon quantity without regard to the facts of the case, should be accepted. True it is that his Honour did not overlook the objective facts, or indeed any other matters relating to penalty, but having started where he did, at a maximum, and then making deductions from it, he did not make, even in a provisional way, an assessment of the sentence called for by the objective facts. It might or might not be appropriate for a trial judge to state such a provisional view. A judge would rarely be in error in not doing so. It is, after all, a provisional position only.

33. A serious fallacy in his Honour's reasoning is that it assumes that any case involving more than 250 grams of heroin is likely to be a worse case than any case involving only 250 grams or less. That cannot be so in the virtually absolute terms in which his Honour puts it. Little imagination is required to envisage a case involving a relatively small quantity of heroin, as being of very great seriousness, for example, supply to create an addiction in an infant. The qualification which his Honour did make of "other things being equal" was not one to which he gave effect, for in adopting his starting point of 15 years he had no regard to the sorts of matters which could have had any equalising effect. The further defect in the reasoning is a related one. Having started with a penalty which would have been appropriate for the worst possible kind of offence of supply involving up to 250 grams of heroin, Hulme J made no attempt to identify the nature of such a case and to make a comparison of the facts of the principal offence with it.

34. For these reasons the appellant's first ground of appeal succeeds.

35. The appellant's next submission invited the Court to reject sequential or two-tiered approaches to sentencing taking as their starting point the maximum penalty available, and to state as a universal rule to the extent that legislation does not otherwise dictate, that a process of instinctive synthesis is the one which sentencing courts should adopt.

36. No universal rules can be stated in those terms. As was pointed out earlier, much turns on what is meant by a "sequential or two-tiered" approach and, likewise, the "process of instinctive synthesis" may wrongly be understood as denying the requirement that a sentencer give reasons for the sentence passed. So, too, identifying "instinctive synthesis" and "transparency" as antonyms in this debate misdescribes the area for debate.

37. In general, a sentencing court will, after weighing all of the relevant factors, reach a conclusion that a particular penalty is the one that should be imposed. As Gaudron, Gummow and Hayne JJ said in Wong[30]:

    "Secondly, and no less importantly, the reasons of the Court of Criminal Appeal suggest a mathematical approach to sentencing in which there are to be 'increment[s]' to, or decrements from, a predetermined range of sentences. That kind of approach, usually referred to as a 'two-stage approach' to sentencing, not only is apt to give rise to error, it is an approach that departs from principle. It should not be adopted.

    It departs from principle because it does not take account of the fact that there are many conflicting and contradictory elements which bear upon sentencing an offender. Attributing a particular weight to some factors, while leaving the significance of all other factors substantially unaltered, may be quite wrong. We say 'may be' quite wrong because the task of the sentencer is to take account of all of the relevant factors and to arrive at a single result which takes due account of them all. That is what is meant by saying that the task is to arrive at an 'instinctive synthesis'. This expression is used, not as might be supposed, to cloak the task of the sentencer in some mystery, but to make plain that the sentencer is called on to reach a single sentence which, in the case of an offence like the one now under discussion, balances many different and conflicting features.

    In R v Thomson[31], Spigelman CJ reviewed the state of the authorities in Australia that deal with the 'two-stage' approach of arriving at a sentence, in which an 'objective' sentence is first determined and then 'adjusted' by some mathematical value given to one or more features of the case, such as a plea of guilty or assistance to authorities. As the reasons in Thomson reveal, the weight of authority in the intermediate appellate courts of this country is clearly against adopting two-stage sentencing and favours the instinctive synthesis approach. In this Court, McHugh and Hayne JJ, in dissenting opinions in AB v The Queen[32] expressed the view that the adoption of a two-stage approach to sentencing was wrong. Kirby J expressed a contrary view. We consider that it is wrong in principle. The nature of the error can be illustrated by the approach adopted by the Court of Criminal Appeal in these matters. Under that approach, the Court takes, for example, the offender's place in the hierarchy and gives that a particular significance in fixing a sentence but gives the sentencer no guidance, whatever, about whether or how that is to have some effect on other elements which either are to be taken into account or may have already been taken into account in fixing the guideline range of sentences. To take another example, to 'discount' a sentence by a nominated amount, on account of a plea of guilty, ignores difficulties of the kind to which Gleeson CJ referred in R v Gallagher[33] when he said that:

    'It must often be the case that an offender's conduct in pleading guilty, his expressions of contrition, his willingness to co-operate with the authorities, and the personal risks to which he thereby exposes himself, will form a complex of inter-related considerations, and an attempt to separate out one or more of those considerations will not only be artificial and contrived, but will also be illogical.'

    So long as a sentencing judge must, or may, take account of all of the circumstances of the offence and the offender, to single out some of those considerations and attribute specific numerical or proportionate value to some features, distorts the already difficult balancing exercise which the judge must perform." (emphasis in original)

38. Following Wong benches of five judges in New South Wales in R v Sharma[34] and R v Whyte[35] and in South Australia in R v Place[36], have sought to state general sentencing principles to be applied in those States. In the first two of these cases the Court of Criminal Appeal of New South Wales endorsed an approach of instinctive synthesis as a general rule but also accepted as a qualification that departure from it may be justified to allow for separate consideration of the objective circumstances of the crime. On occasions intermediate courts of appeal have however refused to find error where a staged approach has been undertaken. In Place[37] the Court of Criminal Appeal of South Australia (Doyle CJ, Prior, Lander, Martin and Gray JJ) although it rejected a staged approach in general, made it clear that a reduction of penalty for a plea of guilty should be identified. This approach, their Honours held, was in conformity with the relevant sentencing legislation of South Australia.

39. Following the decision of this Court in Wong it cannot now be doubted that sentencing courts may not add and subtract item by item from some apparently subliminally derived figure, passages of time in order to fix the time which an offender must serve in prison. That is not to say that in a simple case, in which, for example, the circumstances of the crime have to be weighed against one or a small number of other important matters, indulgence in arithmetical deduction by the sentencing judges should be absolutely forbidden. An invitation to a sentencing judge to engage in a process of "instinctive synthesis", as useful as shorthand terminology may on occasions be, is not desirable if no more is said or understood about what that means. The expression "instinctive synthesis" may then be understood to suggest an arcane process into the mysteries of which only judges can be initiated. The law strongly favours transparency. Accessible reasoning is necessary in the interests of victims, of the parties, appeal courts, and the public. There may be occasions when some indulgence in an arithmetical process will better serve these ends. This case was not however one of them because of the number and complexity of the considerations which had to be weighed by the trial judge.

40. The third ground is that the Court of Criminal Appeal in substance and in fact wrongly imposed a separate sentence of 18 months to 2 years for the further offences. To understand this ground it is necessary to explain the relevant provisions in the Sentencing Act.

41. Division 3 of Pt 3 of the Sentencing Act is concerned with offences other than the particular or principal offence with which a person is charged, and the effect, which in some circumstances an admission of the commission of the former should have upon the penalty to be imposed for the latter. Section 31 of the Sentencing Act defines an additional charge as a further offence. Pursuant to s 32 of the Sentencing Act the prosecutor may file in the court a document (Form 1) specifying other offences with which the offender has been charged but not convicted and that he has indicated he wishes to be taken into account when the principal offence with which he has been charged is dealt with. The relevant requirements of request and the exercise of the prosecutor's discretion were satisfied here. So too the sentencing court in its discretion and on the admission of guilt by the appellant decided to take the further offences into account in dealing with the appellant for the principal offence.

42. Although the appellant did not initially put the matter this way it emerged in argument that his substantial complaint with respect to his third ground of appeal was that the Court of Criminal Appeal ignored the negative direction in s 34(1) of the Sentencing Act which provides as follows:

    "If a court takes a further offence into account under this Division, the court may make such ancillary orders as it could have made had it convicted the offender of the offence when it took the offence into account, but may not impose a separate penalty for the offence."

43. We are not satisfied that the Court of Criminal Appeal did err as contended. For the reasons which we have given the errors of the Court of Criminal Appeal were errors of principle made at the outset, and the effect of referring in terms to an increase in the sentence for the principal offence of between 18 months to 2 years, tended to compound the initial error rather than to constitute a separate error in the application of the Sentencing Act. Just as on occasions, albeit that they may be rare ones, it may not be inappropriate for a sentencing court to adopt an arithmetical approach, it may be useful and certainly not erroneous for a sentencing court to make clear the extent to which the penalty for the principal offence has been increased on account of further offences to which an offender has admitted guilt. Here Hulme J sought to, and in our opinion did make it clear, that the additional period of imprisonment was imposed not as a separate penalty for the further offences but by way of increase of penalty for the principal offence.

44. There was a fourth argument advanced by the appellant: that the sentence of the Court of Criminal Appeal was in any event manifestly excessive and that the sentence of the sentencing judge should be restored. It would need a very exceptional case indeed for this Court to hold that a sentence of an intermediate criminal appellate court of a State was manifestly excessive. This Court is not a sentencing court. It would be most unlikely to be sufficiently aware, for example, of relevant matters such as the prevalence of a particular offence and sentencing patterns in a particular State. This is not a case in which this Court should or could usefully intervene with respect to the duration of sentence and the ground relating to its claimed excessiveness fails.

45. The appellant's appeal must be upheld.

46. The appellant submitted that the sentence of the sentencing judge should be restored. One arguable ground for doing so is that the appeal to the Court of Criminal Appeal was a Crown appeal and that it would be unfair to subject the appellant to a further hearing of it. We do not think that the argument should however be accepted. True it is that in Crown appeals different considerations from those arising on an offender's appeal arise and have to be taken into account. Nonetheless the Crown is entitled to proper consideration of an appeal duly made. That has not happened here. This Court is not, as we have said, in general a sentencing court. We are unable to say whether, having regard to comparable sentences in New South Wales and other relevant matters, the sentence of the sentencing judge is correct or not.

47. We would therefore order that the appeal be upheld, the sentence and orders of the Court of Criminal Appeal of New South Wales be quashed and that the matter be remitted to the Court of Criminal Appeal for disposition of the appeal in accordance with these reasons.

48. McHUGH J. This is an appeal against a sentence imposed by the Court of Criminal Appeal of New South Wales allowing a Crown appeal against the sentence imposed at first instance. The appellant was originally sentenced to two years and six months with a non-parole period of 15 months. The Court of Criminal Appeal re-sentenced him to a term of eight years imprisonment with a non-parole period of four years and six months.

49. The appellant had pleaded guilty to a charge that he knowingly took part in the supply of a prohibited drug, namely heroin, in a commercial quantity - 415 grams. He had further offences taken into account in accordance with the Form 1 process. They included other drug supply offences.

50. In my opinion, the Court of Criminal Appeal made a number of errors in imposing the sentence that it did. They are set out in the joint judgment of the Court. The errors are such that the decision of the Court of Criminal Appeal must be set aside. If there were no more to the case, I would need to do no more than agree with the joint judgment. But there is more to the case than whether or not the Court of Criminal Appeal erred in imposing the sentence that it did. The appellant was granted special leave to appeal in this case because he contended that the key question in the case was whether "two-tier sentencing" in contrast to "instinctive synthesis" is the correct approach to sentencing. In this case, the Court of Criminal Appeal applied the "two-tier" approach. It erred in doing so.

51. By two-tier sentencing, I mean the method of sentencing by which a judge first determines a sentence by reference to the "objective circumstances" of the case. This is the first tier of the process. The judge then increases or reduces this hypothetical sentence incrementally or decrementally by reference to other factors, usually, but not always, personal to the accused. This is the second tier. By instinctive synthesis, I mean the method of sentencing by which the judge identifies all the factors that are relevant to the sentence, discusses their significance and then makes a value judgment as to what is the appropriate sentence given all the factors of the case. Only at the end of the process does the judge determine the sentence.

52. The two-tier sentencer contends that using the instinctive synthesis is inimical to the judicial process and is an exercise of arbitrary judicial power, unchecked by the giving of reasons. The two-tier sentencer claims, as Hulme J did in this case, that, where the sentence is the result of an instinctive synthesis, it makes one "wonder whether figures have not just been plucked out of the air"[38]. The instinctive synthesiser, on the other hand, contends that the two-tier sentencer mistakes an illusion of exactitude for the reality of sentencing because there is no method of sequential arithmetical reasoning that produces the correct sentence for any case. A sentence can only be the product of human judgment, based on all the facts of the case, the judge's experience, the data derived from comparable sentences and the guidelines and principles authoritatively laid down in statutes and authoritative judgments. The instinctive synthesiser asserts that sentencing is not an exercise in linear reasoning because the result of each step in the process is not the logical foundation for the next step in the process. Nor in practice can it be an exercise in multiple regression where one starts with particular coefficients and adds to or subtracts from their result by changing the weighting of each variable as new variables are added to the process. The circumstances of criminal cases are so various that they cannot be the subject of mathematical equations. Sociological variables do not easily lend themselves to mathematization. Hence, when judges embark on a process that seeks to adjust incrementally or decrementally a hypothetical sentence, "they but illustrate the way in which the human mind tries, and vainly tries, to give to a particular subject matter a higher degree of definition than it will admit", as Lord Porter said[39] in another context.

53. In AB v The Queen[40], I gave my reasons for preferring the instinctive synthesis approach. In my view, the judge who purports to compile a benchmark sentence as a starting point inevitably gives undue - even decisive - weight to some only of the factors in the case. Furthermore, the judge falls into the error of determining that notional sentence by reference to a hypothetical crime derived from some only of the circumstances of the case. Instead of sentencing this accused for his or her criminality, the judge sentences the person for another crime and adjusts the notional sentence by reference to factors that are additional to the objective circumstances. Indeed, there are some offences - manslaughter is an example - where an attempt to fix a first-tier sentence by reference to the objective circumstances is meaningless. How can a judge possibly fix a first-tier or any sentence for the mother who has killed her newborn baby without taking into account her personal circumstances?

54. Moreover, by concentrating on the objective circumstances of a crime, the judge is giving effect, and ultimately greater weight, to the retributive or deterrent theory of sentencing. Indeed, the judgment of the Court of Criminal Appeal in this case makes it clear that the Court thought that the issues of retribution and deterrence were the dominant issues in the case. Consciously or unconsciously, the judge who commences with a notional sentence downplays the importance of mitigation, reformation and rehabilitation in the sentencing process. Cognitive psychology has long emphasised the difficulty that the human mind has in giving correct weightings to each of a number of variables. In particular, people frequently fail to distinguish between the strength of evidence and its relative weight in determining the outcome or prediction. As Griffin and Tversky have pointed out[41]:

    "The extensive experimental literature on judgment under uncertainty indicates that people do not combine strength and weight in accord with the rules of probability and statistics."

    The tendency of the mind is to seize on one or two variables - usually those with which the decision-maker is most familiar or which seem most cogent - and give that variable or those variables undue weight. Overconfidence - but sometimes underconfidence - in the significance of factors or the accuracy of the assessment is very common. The tendency to err must increase when particular circumstances are selected as the starting point for the decision and further factors are allowed to modify the starting point.

55. One fact that critics of the instinctive synthesis approach do not face up to - assuming they are aware of it - is that the first tier of the two-tier approach - unless it is the maximum sentence - is itself derived by an instinctive synthesis of the "objective circumstances" of the case. Or on another view of the two-tier approach, the first-tier sentence is the product of a value judgment that is proportionate to the offence. But as the Victorian Court of Criminal Appeal said in R v Young[42]:

    "What is a sentence proportionate to an offence is a matter of discretion and there must in most cases be a range of sentences open to a sentencing judge which are proportionate to the offence. There cannot be said to be a sentence which is the proportionate sentence ... Thus to attempt to fix a proportionate sentence before fixing the sentence to be imposed will only multiply the possibilities of error. Upon what facts is the proportionate sentence to be fixed?" (emphasis in original)

56. Analysing the process involved in two-tier sentencing reveals that its appearance of objectivity and unfolding reason is illusory. Whether the starting point is a sentence derived from the objective circumstances or a sentence proportionate to the offence, the correctness of the sentence always depends on the correctness of the value judgment involved in assessing the first-tier sentence. But even if the judge can correctly assess the first-tier sentence, the judge must still correctly assess the quantum of the increment or decrement for each factor in the process. With great respect to those who think the contrary, it would require a judge to have the statistical genius and mental agility of a Carl Friedrich Gauss[43] to arrive at the correct sentence using these methods. As Gaudron, Gummow and Hayne JJ pointed out in Wong v The Queen[44], mathematical increments and decrements to some pre-determined notional sentence are "apt to give rise to error".

57. The Court of Criminal Appeal's judgment in the present case is a nice illustration of this tendency. In giving the leading judgment, Hulme J said "that the maximum [15 years] prescribed for the supply of 250 g is not too high a starting point."[45] His Honour then said that only five topics removed the appellant's sentence "from a worst case category - his role, his plea, the finding of contrition, his addiction, and matters which fall within the topic of rehabilitation."[46] The learned judge then went on to say[47]:

    "In light of the matters referred to in this paragraph, I would reduce my 15 year starting point by about one-third. A number of factors lead me to the view that the reduction on this account should not be greater. These include the sorts of considerations spoken of in Le Cerf[48]. They include also my view that the severity of sentences is not simply proportionate to length. They include the nature of [Markarian's] activities and the fact that they extended over a much longer period than that during which a courier is normally involved. The conclusion derives some support also from the relativity between the maximum sentences available for importing heroin and the pattern of sentences imposed on couriers involved in the importation of quantities in the top part of trafficable quantities of that drug."

58. This reasoning process was wholly dependent on a series of value judgments and quantification of intangibles. Ironically, his Honour, having criticised[49] the primary judge for not advancing reasons for his starting point of 3.5 years, justified the maximum as the starting point by simply referring to "the totality of the statutory provisions and the principles for which I have cited Veen v The Queen [No 2][50] and R v Peel[51]"[52]. But with great respect, this is a less than illuminating revelation of reasons. What it shows is that almost invariably the major premise of the two-tiered sentence is a value judgment based on the judge's instinct or intuition. His Honour then differentiated the present case from the "worst case category"[53] by reference to a number of identified factors. But how does a judge decide the "worst case category", except by judicial imagination? The next step was to assert that these factors required a one-third reduction from the maximum sentence attributable to the "worst case category". But his Honour did not use the maximum sentence for the offence with which the appellant was charged. Rather, he used the maximum for the offence of supplying less than 250 grams of heroin, apparently on the basis that "Parliament cannot have intended that, other things being equal, the penalty for supplying more than 250 g should be less than for supplying that quantity."[54] His Honour did not provide any other justification for this approach. Moreover, is not this one-third reduction a figure "plucked out of the air"? Hulme J then applied a 25% discount for the appellant's guilty plea in respect of all the offences making "it clear that it covers all credit [Markarian] may be entitled to in consequence of the plea."[55] Again the 25% figure - which the trial judge had also applied - is an arbitrary figure even though it is a discount for pleading guilty that is frequently applied by judges.

59. The result at this stage was that the 15 years starting point had been reduced by one-third to 10 years and then reduced by 25%. This would have resulted in a sentence of 7.5 years. His Honour then said[56]:

    "Operating in the other direction are the offences on the Form 1. I have said sufficient to indicate my view about them save and except that principles of totality have also to be taken into account. On account of the matters on the Form 1, particularly the first and second of these, I would increase the sentence otherwise appropriate by between 18 months and two years."

60. This statement would seem to suggest that the sentence of 7.5 years - a figure which Hulme J never specifically mentioned but which was already determined before this statement - would increase to either 9 or 9.5 years. But this is not what occurred, and I do not understand why it did not occur. Instead, Hulme J arrived at a sentence of eight years. Nothing in the remaining reasons reveals why the sentence was not increased beyond eight years. After referring to the appellant's prospects of successful rehabilitation, his Honour said[57]:

    "As [his] efforts to date in this regard would seem to have been appreciably more than token, I am disposed to reduce the sentence to a limited extent, but because the value of his efforts depends so much on their success, it seems to me that the topic should be reflected more in the non-parole period than in the head sentence."

61. His Honour said that there was nothing in Markarian's evidence or the Pre-Sentence Report "which has any appreciable significance."[58] He referred to an affidavit of the appellant and said it could not have "any material impact on the question of what should be a proper sentence."[59] His Honour then said[60]:

    "Thus, on the basis of the matter to which I have referred, I would impose on [Markarian] a sentence of eight years. In proposing that figure I make it clear that in arriving at it I have taken account of the fact of the double jeopardy to which [he] has been subjected and selected a sentence which is the lowest that could reasonably be imposed."

62. Given his Honour's remarks about rehabilitation, Markarian's evidence, the Pre-Sentence Report and the affidavit, it seems unlikely that his Honour gave the appellant a credit of up to two years for these factors. But perhaps he did. Or it may be that it was taken into account earlier in arriving at the starting point of the sentencing process. Or maybe his Honour decided intuitively that a sentence of 9 or 9.5 years was too high.

63. His Honour concluded by saying that neither the principles of parity nor the sentencing statistics "kept by the Judicial Commission"[61] required a sentence different from the eight years his Honour proposed. His Honour acknowledged that the sentence was comparatively high and was imposed on a Crown appeal but held that in all the circumstances it was the lowest appropriate sentence that could be imposed. This suggests that the two-tier approach led his Honour to believe that a sentence of up to 12 years may have been appropriate.

64. No doubt the process in which Hulme J engaged revealed his erroneous quantification of various elements in the sentence as well as the invalidity of his major premise or starting point. But it also revealed the arbitrariness of the two-tier approach and its almost exponential capacity for error. It showed that the criticism that his Honour directed at the instinctive synthesis approach applied to each stage of his own reasoning process. It is no answer to the criticism that the two-tier approach creates error to say that, because it reveals the error, it permits an appellate court to correct the error. The need for appellate intervention arises only because the two-tier approach is inherently susceptible to error. Appellate counsel are unlikely to be short of material to attack the reasoning process of judges who use the two-tier approach. Appeals are inevitable, and likely to succeed. Ironically, sentences imposed by using the two-tier method are likely to be upheld only by appellate courts declaring that, given the circumstances, there has been no miscarriage of justice because the sentence imposed was within the appropriate range.

65. Unfortunately, discretionary sentencing is not capable of mathematical precision or, for that matter, approximation. At best, experienced judges will agree on a range of sentences that reasonably fit all the circumstances of the case. There is no magic number for any particular crime when a discretionary sentence has to be imposed. What Jordan CJ said in R v Geddes[62] about the reality of the sentencing process has never been bettered and probably has never been equalled. With the passage of time, it is no longer cited as frequently as it once was. But the whole judgment repays careful study. I make no apology for setting out the crucial passage, lengthy though it is:

    "This throws one back upon a preliminary question as to the general principles upon which punishment should be meted out to offenders. In the nature of things there is no precise measure, except in the few cases in which the law prescribes one penalty and one penalty only. In all others, the judge must, of necessity, be guided by the facts proved in evidence in the particular case. The maximum penalty may, in some cases, afford some slight assistance, as providing some guide to the relative seriousness with which the offence is regarded in the community; but in many cases, and the present is one of them, it affords none. The function of the criminal law being the protection of the community from crime, the judge should impose such punishment as, having regard to all the proved circumstances of the particular case, seems, at the same time, to accord with the general moral sense of the community in relation to such a crime committed in such circumstances, and to be likely to be a sufficient deterrent both to the prisoner and to others. When the facts are such as to incline the judge to leniency, the prisoner's record may be a strong factor in inducing him to act, or not to act, upon this inclination. Considerations as broad as these are, however, of little or no value in any given case. It is obviously a class of problem in solving which it is easier to see when a wrong principle has been applied than to lay down rules for solving particular cases, and in which the only golden rule is that there is no golden rule.

    The position of the judge is analogous to that of a civil jury who are called upon to award damages for a breach of contract, or a tort, in relation to goods which have no market value, and for the assessment of the value of which no generally accepted measure exists. The jury must do the best they can; and so must the judge. In applying considerations as general as these, it is necessarily not often that it can be said, with reasonable confidence, that the sentence imposed was wrong." (emphasis added)

66. This passage is not a testament of despair but a perceptive understanding of the reality of the sentencing process by one of the greatest judges that the common law system of justice has produced. It recognises that the judge must weigh all the circumstances and make a judgment as to what is the appropriate sentence. In R v Williscroft[63], the Full Court of the Supreme Court of Victoria referred to this value judgment as an "instinctive synthesis of all the various aspects involved in the punitive process." This was a candid recognition of the fact that in the end sentencing depends on the judge's assessment of what is the correct sentence. There is no objectively correct sentence, only a range of sentences that the majority of experienced judges would agree applied to the case. The only novelty in Williscroft was the description that it gave to the sentencing process.

67. After I came to the Bar in 1961, judges in New South Wales, to the best of my recollection, always followed the approach referred to in Williscroft. Only in recent times has there been any attempt to move away from it. Having regard to the remarks of Jordan CJ in Geddes, it seems almost certain that New South Wales judges were applying the Williscroft approach long before 1936. In 1961 and for some years afterwards, the most cited cases in this area of law in New South Wales were Geddes[64], Goodrich[65], Cooke[66] and Herring[67]. Those four cases reflected the general principles applied and the approach taken to the sentencing process. The reasoning in those cases followed the Williscroft approach. There were, of course, other cases in New South Wales, England and other States that required particular matters to be taken or not taken into account. But these four cases illustrated the proper approach in all cases. The judgment of Street CJ in Herring[68], where the Court of Criminal Appeal allowed a Crown appeal and re-sentenced the prisoner, is a good and concise illustration of the way that New South Wales judges approached the sentencing issue. After reviewing the facts and referring to the principles concerning retribution, reformation, protection of the public and mitigation, Street CJ concluded[69]:

    "I view this offence most seriously. I do not think that it calls for the maximum penalty, but it certainly calls for nothing light. If the prisoner had been older than he is, or if he had had any previous record indicating that he was not entitled to expect leniency from this Court, I would have imposed a heavier sentence than that which I think is the proper one in the present case. But giving full weight to everything that can be said in favour of the prisoner, I think the proper sentence to impose to mark this Court's view of the seriousness of the crime, and to let other wrongdoers know the retribution which will fall upon them if they commit similar crimes, is one of five years' penal servitude ..."

68. Nothing in any High Court judgment before or since Williscroft throws any doubt on the approach in that case. In particular, there is not a line in the joint judgment in Veen v The Queen [No 2][70] that supports the two-tier approach to sentencing. As the Victorian Court of Criminal Appeal pointed out in R v Young[71], the critical "passage demonstrates in clear terms the same approach to the fixing of an appropriate sentence as this court adopted in Williscroft's Case." And as the Court also said in Young[72]:

    "There is also much in the majority judgment in Veen (No 2) which shows that the High Court simply did not have in mind that a sentencer might, let alone should, proceed to arrive at the sentence to be imposed by a staged or structured approach."

    Both Veen [No 2] and the earlier Veen[73] case were concerned with the issue of whether the penalty was proportionate to the facts of the case.

69. The principle of proportionality is one of the fundamental principles of sentencing law. It is difficult - maybe impossible - to reconcile that principle with the two-tier approach to sentencing. The principle of proportionality requires the judge to make a judgment concerning the relationship of the penalty to the facts. This is a value judgment, based on experience and instinct, derived after taking into account all the facts and circumstances of the case. The existence of the proportionality principle makes one wonder whether, despite appearances, two-tier sentencers exist. At the end of the process, the two-tier sentencer must ask whether the result of the additions and subtractions from the objectively determined sentence is proportionate to the accused's offence. What happens if the judge concludes that the result is not proportionate to the offence? It would be almost a miracle if it was. If the judge tinkers with the quantum of each component in the sentence to achieve a result compatible with the concept of proportionality, the two-tier structure is meaningless, if not a charade.

70. Whether or not the two-tier approach to sentencing does exist in practice, the common law should not accept it as superior to the method that in Williscroft the Court called the instinctive synthesis. Nothing in the judgments that have used it suggests that the two-tier approach will produce sentences that are more acceptable or better than the sentences produced by the instinctive synthesis method. The judgment of the Court of Criminal Appeal in this case is an example of its inherent tendency to error. I think it very likely that a judge with the experience of Hulme J would have imposed a proportionate sentence if he had used the instinctive synthesis approach instead of the approach he followed. But unfortunately the learned judge engaged in an arithmetical exercise instead of applying his judgment based on his considerable judicial experience with the result that he fell into error.

71. The belief that two-tier sentencing is the preferable method is principally based on the idea that it promotes transparency of sentencing. Certainly, it shows a series of numbers. But they are more likely than not to be erroneous numbers. Each time the judge adds or subtracts another number the chance of ultimate error increases exponentially. In so far as its proponents claim that two-tier sentencing also promotes predictability, they mistake the illusion for the reality. Its proponents also contend that it makes sentencing more scientific. But if two-tier sentencing is science, its results, as in this case, suggest it is junk science. Belief in the advantages of two-tier sentencing is reminiscent of the once popular belief that judges do not make law. Like that belief, it belongs in Lord Reid's fairytales. There is no Aladdin's Cave of accurate sentencing methodology, the door to which can be opened by chanting the magic words, "two-tier sentencing". There is only human judgment based on all the facts of the case, the judge's experience, the data derived from comparable sentences and the guidelines and principles authoritatively laid down in statutes and authoritative judgments.

72. The flaws in the two-tier method do not mean that the instinctive synthesis approach is perfect. Far from it. Any assessment, based on indeterminate standards and human judgment - whether it is negligence, damages or sentences - is unsatisfactory. That is why I have always preferred the use of rules and principles to standards. And the instinctive synthesis method is open to the criticism that, in arriving at the sentence, the judge has unconsciously over-emphasised or under-emphasised the weight to be given to various factors in the synthesis. But unless we adopt fixed sentences or provisions similar to the United States Federal Sentencing Guidelines with their requirements that courts must impose a sentence of the kind and within the range mandated[74], it is the best we can do.

73. Critics of the instinctive synthesis method place too much emphasis on the "instinct" and too little on the "synthesis". The use of the word "synthesis" in the context of sentencing identifies the very last part of the process. It recognises that, where a variety of considerations, often tending in opposing directions, operate in the context of a statutory maximum, there must finally be a quantification of the sentence to be imposed. There must be a synthesising of the relevant factors. In that process, greater and lesser weight will be allocated to some factors depending on their relevance to the person convicted and his or her crime. Ultimately, community and legal values are translated into a number of years, months and days. That process must involve an instinctive judgment. As Mason CJ, Brennan, Dawson and Toohey JJ said in their joint judgment in Veen [No 2][75]:

    "[S]entencing is not a purely logical exercise, and the troublesome nature of the sentencing discretion arises in large measure from unavoidable difficulty in giving weight to each of the purposes of punishment. The purposes of criminal punishment are various: protection of society, deterrence of the offender and of others who might be tempted to offend, retribution and reform. The purposes overlap and none of them can be considered in isolation from the others when determining what is an appropriate sentence in a particular case. They are guideposts to the appropriate sentence but sometimes they point in different directions." (emphasis added)

74. Nor is the instinctive synthesis approach inconsistent with awarding a discount for some factor, provided that discount relates to a purpose distinct from a sentencing purpose. The distinction between permissible and impermissible quantification of "discounts" on a sentence will usually be found in whether the quantification relates to a sentencing purpose rather than some other purpose. So, the quantification of the discount commonly applied for an early plea of guilty or assistance to authorities is offered as an incentive for specific outcomes in the administration of criminal justice and is not related to sentencing purposes. The non-sentencing purpose of the discount for an early guilty plea or assistance is demonstrated by the fact that offenders are ordinarily entitled to additional mitigation for any remorse or contrition demonstrated with the plea or assistance, aside from the discount for willingness to facilitate the course of justice[76]. That said, I think the use of discounts should be reserved for only one - maybe two - factors in a particular sentence that serve some goal other than a sentencing goal.

75. In this case, Hulme J, having commenced from an erroneous premise, applied a fractional reduction that reflected the lesser gravity of the appellant's role in the offences, the lengthy period of offending, and some reference to proportionality with sentences for related offences. In so doing, Hulme J was selecting matters related to retribution and general and specific deterrence. His Honour was giving a mathematical value to these purposes separated from the other purposes that were to be synthesised in the sentencing outcome. These were not appropriate matters for separate quantification and caused his Honour to fall into error.

76. One reason why the idea of instinctive synthesis is apparently abhorrent to lawyers who value predictability and transparency in sentencing is that they see the instinct of a sentencing judge as entirely subjective, personal, arbitrary and unconfined. In fact, although a sentencing judge does ultimately select a number, it is not from thin air that the judge selects it. The judicial air is thick with trends, statistics, appellate guidance and, often enough these days, statutory guidance.

77. First, the sentencing judge almost never imposes a sentence for an offence that has been committed for the first time. A sentencing judge may have seen dozens or scores of such cases and develops, through experience, a sense of the relative gravity of offences and the relative circumstances of offenders that dictate the weighting of different factors in the sentencing process. The need to give greater weight to general or specific deterrence in response to crime trends is one factor to which a sentencing judge has special sensitivity. A sentencing judge also has the benefit of collegiate knowledge, both formally through reading the judgments of other judges and informally through interaction with other judges.

78. No one suggests that the judicial robe carries in its seams the wisdom of Solomon, but judicial experience in sentencing is a skill to be respected by the community and other judges. Repeated exercise in synthesising sentencing factors can only hone the instinct required to translate such factors into just numerical outcomes. That experience, combined with the special advantages of receiving sentencing material, including oral material, first hand, are the two most important reasons why appellate courts, and especially an ultimate appellate court which is national rather than local to the sentencing jurisdiction, must exercise restraint in reviewing sentencing decisions, especially on the basis of manifest excess or leniency.

79. A further source of information about the sentences imposed by other judges is the sentencing statistics produced by (in New South Wales) the Judicial Commission. Hulme J referred to these statistics towards the end of his judgment in this case[77]. It is surprising that they did not cause the Court of Criminal Appeal to see that the sentence of eight years that it was imposing was disproportionate. Those statistics showed that the Court was imposing a sentence as high as any that had been imposed during a six year period dealing with 22 cases concerning the same offence, despite the subordinate role played by this offender and the context of a Crown appeal. The failure of the Court to act on those statistics suggests that its belief in the "logic" of its numbers caused it to overlook the significance of the statistics. If so, it shows the dangers lurking in an approach that concentrates on numerical components.

80. Second, a judge is sensitive to legislative trends. A change in the maximum penalty for an offence or in the elements of an offence may indicate a shift in the values to be applied when sentencing for that offence. In New South Wales there is also a statutory system of guideline judgments and standard minimum non-parole periods that give more specific guidance in common offences and operate as a starting point from which departure is intended to be the exception or at least require explanation. In recent times, both methods have been used to increase the prevailing median sentence for particular classes of offences. That does not mean that the judge must start with a specific number but knowledge of the median or the extent of the range guides the judicial "instinct".

81. Third, a sentencing judge always knows that the sentence imposed is subject to judicial review by an appellate court. Whether or not that review takes place, he or she is conscious that the sentencing discretion and the reasons for arriving at a particular sentence will be considered by the advisers to the Crown and the offender. Error will be the subject of appeal. To avoid appealable error, a judge pays close attention to the guidance provided by appellate courts as to the impermissible paths of reasoning and the permissible factors which will be relevant to the sentencing process in a particular case.

82. Fourth, the role of open justice is also important. A judge's sentence and reasons are usually exposed to public scrutiny through publication or media reporting. Public responses to sentencing, although not entitled to influence any particular case, have a legitimate impact on the democratic legislative process. Judges are aware that, if they consistently impose sentences that are too lenient or too severe, they risk undermining public confidence in the administration of justice and invite legislative interference in the exercise of judicial discretion. For the sake of criminal justice generally, judges attempt to impose sentences that accord with legitimate community expectations.

83. Finally, in Veen [No 2][78], as I have indicated, this Court affirmed that the ultimate control on the judicial sentencing discretion is the requirement that the sentence be proportionate to the gravity of the offence committed. In pursuit of other sentencing purposes, a judge may not impose a sentence that is greater than is warranted by the objective circumstances of the crime. Both proportionality and consistency commonly operate as final checks on a sentence proposed by a judge. They guard against hidden errors in the process, the kind later identified on appeal as manifest excess or leniency in accordance with the principles in House v The King[79].

84. The acceptance of the role of instinctive synthesis in the judicial sentencing process is not opposed to the concern for predictability and consistency in sentencing that underpins the rule of law and public confidence in the administration of criminal justice. The synthesising task is conducted after a full and transparent articulation of the relevant considerations including an indication of the relative weight to be given to those considerations in the circumstances of the particular case. The instinctive synthesis approach does not prevent the use of adjectives or adverbs or indications that this or these factors makes or make the case more or less serious than other cases or are the critical features of the case. And judicial instinct does not operate in a vacuum of random selection. On the contrary, instinctive synthesis involves the exercise of a discretion controlled by judicial practice, appellate review, legislative indicators and public opinion. Statute, legal principle and community values all confine the scope in which instinct may operate. The judicial wisdom involved in the instinctive synthesis approach is therefore likely to lead to better outcomes than the pseudo-science of two-tier sentencing. At all events, I am not satisfied that two-tier sentencing is a better method or process than the instinctive synthesis method that has been the traditional approach of common law judges.

    Order

85. The appeal must be allowed.

86. KIRBY J. This is an appeal from a judgment of the Court of Criminal Appeal of New South Wales[80]. That judgment upheld a prosecution appeal to that Court from a sentence which Hosking DCJ imposed on Mr Anthony Markarian (the appellant) in the District Court of New South Wales. As a result of upholding the appeal, the Court of Criminal Appeal set aside the sentence of two years six months imprisonment from 18 July 2002 (with a non-parole period of fifteen months). It substituted a sentence of eight years imprisonment (with a non-parole period of four years six months).

87. The appeal to this Court has concerned the appellant's submission that, in disturbing the sentence imposed on him in the District Court, the court below erred and that the substituted sentence (and the reasons given for it) disclosed either specific or imputed error[81]. The appellant asked this Court to restore the sentence imposed on him by Hosking DCJ or, at the least, to return the matter to the Court of Criminal Appeal for reconsideration of the prosecution's appeal to that Court, freed from the errors complained of.

    The facts, legislation, issues and disposition

88. The facts and related sentences: Most of the background facts are stated in the reasons of Gleeson CJ, Gummow, Hayne and Callinan JJ ("the joint reasons")[82]. It is necessary, however, to appreciate that the sentencing of the appellant in the District Court occurred as one of a series of related judicial acts affecting four prisoners who were severally involved in an illicit business, conducted by Mr Vincent Caccamo, involving the handling and sale of illegal drugs.

89. On 30 May 2002, Mr Caccamo himself was sentenced by Shillington DCJ to eight years imprisonment (with a non-parole period of five years). Allowance was made in his case for his assistance to the authorities and for the consequent requirement that Mr Caccamo would serve his sentence in protective custody. But for those facts, Shillington DCJ said that he would have sentenced Mr Caccamo to 15 years imprisonment. On 14 June 2002, Mr Chung was sentenced by Knight DCJ to three years imprisonment, with a non-parole period of two years, to be served by way of periodic detention. Unlike the appellant, he had no criminal record. However, like him, Mr Chung was a driver for Mr Caccamo, although employed less frequently than the appellant had been. The appellant was sentenced on 18 July 2002 by Hosking DCJ. As stated, his sentence was two years six months imprisonment (with a non-parole period of fifteen months). There followed the sentencing of a fourth prisoner, Mr Barta, who was sentenced by Hosking DCJ on 25 July 2002[83]. He was the person who had supplied Mr Caccamo with heroin. His sentence is undisclosed.

90. It is proper to infer that, in sentencing the named prisoners for their respective offences (whilst also taking into account disclosed further offences), the judges of the District Court would have kept in mind, in a general way, the respective roles of the offenders in the criminal enterprise which had Mr Caccamo at its centre, having regard to the comparative criminality of each of them. This notwithstanding, the sentence imposed on the appellant by the Court of Criminal Appeal resulted in an obvious and serious disturbance of the relativities of the punishments imposed on the respective offenders. It left the appellant, a heroin-addicted chauffeur charged only with knowing participation in the supply of the drug by Mr Caccamo, with a final sentence identical to that imposed on Mr Caccamo, the ring-leader and mastermind of the criminal business. The appellant's non-parole period, moreover, was just six months short of that fixed in Mr Caccamo's case. When such a result is arrived at, alarm bells begin to ring.

91. The applicable legislation: The joint reasons identify the three statutes of the New South Wales Parliament that afford the legislative context in which the task of sentencing was performed both by the primary judge and by the Court of Criminal Appeal.

92. The first was the Drug Misuse and Trafficking Act 1985 (NSW), s 33(2), which fixed the maximum sentence that Parliament had provided for the offence with which the appellant was charged[84]. Also contained in the joint reasons are references to the Crimes (Sentencing Procedure) Act 1999 (NSW) ("the Sentencing Act")[85]. Finally, it is proper, as the joint reasons state, to keep in mind the terms of the Criminal Appeal Act 1912 (NSW), s 5D of which afforded the Court of Criminal Appeal its jurisdiction and powers with respect to the prosecution appeal against the sentence imposed on the appellant. To enliven the power to re-sentence the appellant, it was necessary for the appellate court to be satisfied of error, either specific or imputed, on the part of the sentencing judge[86].

93. I agree with the joint reasons that no task of sentencing, at trial or on appeal, could be accurately carried out without proper attention to any statutes affecting the maximum penalty fixed by Parliament for the worst possible case; the procedure to be followed where a judicial duty to sentence, or re-sentence, the appellant was enlivened; and the precondition to be observed where appellate powers of re-sentencing were invoked. In this respect, the statutory requirements applicable to sentencing constituted the starting points for the judicial task. That is because, in a case like the present, whatever has been said or written by judges about that task must take second place to the requirements of legislation as part of the written law. So long as that law is constitutionally valid, it has the imputed authority of democratic credentials and must be obeyed[87].

94. The sentence and re-sentence: The joint reasons explain the approach taken by Hosking DCJ to sentencing the appellant[88] and the explanations of Hulme J, who gave the reasons of the Court of Criminal Appeal, for disturbing that sentence and arriving at a new, increased sentence[89]. Hulme J is a judge with long experience in criminal trials and sentencing. His detailed and thorough treatment of the many issues presented by the appellant's case bears witness to his close attention to the issues and to his endeavour to identify and pay regard to all of the considerations that he regarded as relevant and applicable to the task in hand.

95. It will be evident from past expositions of my own[90], that I share with Hulme J (and inferentially with Heydon JA and Carruthers AJ in the Court of Criminal Appeal, who agreed with Hulme J's reasons) a discomfort with talk about determining sentences "by instinctive synthesis"[91]. I would point out at the outset of these reasons that Hulme J's disquiet about so-called "instinctive synthesis" derived, as he explains, from his experience as a judge, both at trial and on appeal, in seeking, as conscientiously as he did in this case, to reach an outcome sustained by a process of reasoning more satisfying and conformable to the rule of law than a so-called judicial "instinctive synthesis". The danger of that language, as Hulme J correctly explained, is that it can lead all too easily to figures "just ... plucked out of the air"[92].

96. Identifying elements in a sentence: I shall return to this point. But I cannot leave it without observing that it is the very process, accepted by Hulme J as part of his judicial function in this case, of explaining in detail and with proper care the manner of arriving at the sentence that he believed should be substituted for that imposed at first instance, that opens up the effective remedy which the appellant now invokes in this appeal.

97. Had Hulme J and his colleagues done little more than to suggest that the sentence imposed by Hosking DCJ struck them as "manifestly inadequate" and thus exhibiting imputed error calling for a judicial "synthesis" attributed to "instinct" or a sequential consideration of relevant but unidentified factors, the appellant's appeal to this Court would have been bound to fail. Indeed, the appellant would almost certainly have been refused special leave to appeal[93]. As the joint reasons accept, the gateway to this Court in appeals expressed in terms of manifest excess (or inadequacy) of sentence is almost always barred and locked, and the key is rarely found[94].

98. There is, then, an irony in the fact that it is the very attention of Hulme J and his colleagues to transparency in the re-sentencing process that has made it feasible for the appellant to attract the attention of this Court and now to secure relief from the judgment of the Court of Criminal Appeal. Error of judicial reasoning is not made more palatable for prisoners (or anyone else) who suffer as a consequence because the judges concerned are encouraged by appellate courts to submerge the true steps taken by their minds beneath talk of "instinctive", "intuitive" or unspecified "sequential" approaches to an outcome declared but inadequately explained.

    Extent of agreement with the joint reasons

99. The omission to express error: Having said this, and having allowed full credit for the exposure by the Court of Criminal Appeal of its reasoning - available for the appellant, the respondent, the community and this Court - I have nevertheless come to the conclusion that its reasoning exhibits error. The errors must be corrected. For the most part, I agree with the conclusions expressed in the joint reasons. I agree with the orders favoured there.

100. Like my colleagues, I am prepared to assume that, although it did not expressly say so, the Court of Criminal Appeal found error on the part of the sentencing judge sufficient to warrant the disturbance of the sentence that he had imposed on the appellant[95]. It is important, once again, to emphasise the point made by this Court in Lowndes v The Queen[96] and in other cases concerned with the exercise of appellate powers[97]. An explicit finding of error by appellate judges is not a mere technicality. It is the precondition to the authority which the appellate court enjoys under the law to disturb the conclusions of the trial judge, manifested in that judge's orders.

101. What is involved in this rule is not simply professional respect for the trial judge, still less for a formula of words. It is a salutary reminder to the appellate court of the advantages that the trial judge enjoys; the impossibility of expressing all of the considerations leading to an outcome in judicial reasons; and the special difficulty of doing so where the outcome involves (as sentencing does) discretionary and quasi-discretionary considerations of judgment. To pause at the end of the analysis of criticisms of the reasons of a trial judge, and to express clearly the appellate court's satisfaction that error has been established, is a useful reminder to the appellate court that its function is different from that of the trial judge. It is so even if, once error is shown, the appellate court enjoys its own separate power to substitute the orders that ought to have been made at trial. Whilst this step was not expressly taken in the instant case, I am satisfied that it was inherent in the reasoning of Hulme J. By its very detail, that reasoning displays what his Honour considered to be the errors of the sentencing judge, and what he regarded as necessary to a proper sentence.

102. The omission to apply the statute: I have more concern about the omission of the sentencing judge and of the Court of Criminal Appeal to pay regard to the requirements of s 21A of the Sentencing Act[98]. As is now common ground, that section applied to this case[99]. It is a rudimentary error in the exercise of a sentencing discretion (or of the discretion enlivened by the appellate re-sentencing of an offender) for the decision-maker to fail to take into account a relevant consideration[100]. It is clearly relevant for a judge engaged in sentencing, or re-sentencing, to pay regard to an applicable provision of the written law, such as the Sentencing Act, made by Parliament to apply to such a case. Statute law, having the higher authority of Parliament, cannot be waived by parties simply because they are ignorant of it or because they choose not to argue it although it is applicable. Once such an omission comes to light in proceedings that are still current within the Judicature, judges, certainly when they are on notice of such provisions, are under a constitutional duty to obey them and give them effect.

103. Although it seems that s 21A of the Sentencing Act was not called to the attention either of the sentencing judge or of the Court of Criminal Appeal, its terms have been brought to our notice, as has the fact that it was not given consideration below. What to do? For my own part, I would not disregard the omission as irrelevant to the proper exercise of the sentencing discretion in the matter before this Court as do the joint reasons[101]. There, reliance is placed on assumptions and the lack of contrary submissions.

104. For me, it is sufficient to say that the omission of the Court of Criminal Appeal to pay specific regard to the aggravating, mitigating and other factors in sentencing elaborated in s 21A of the Sentencing Act can be passed by in this appeal for the practical reason that, in the result of the appeal, for other reasons, the entire matter must be reconsidered by the Court of Criminal Appeal. The sentence is not, as such, invalidated by the omission to comply with the Sentencing Act. It continues to govern the appellant's case[102].

105. Should it become necessary to re-sentence the appellant (or to test the validity of the sentence imposed by Hosking DCJ) on reconsideration by the Court of Criminal Appeal, that can be done by that Court with due attention to the provisions of s 21A. On the face of things, however, the failure to attend to the section was an oversight that cannot, without argument, be said to have been irrelevant to the process of re-sentencing which the Court of Criminal Appeal felt was enlivened by the errors that it found in the sentence imposed by Hosking DCJ. If the statute applied (as is agreed) the duty of the court engaged in sentencing was to "take into account" all of the specified factors and that was clearly a duty binding in law. It was a duty unfulfilled.

106. The reasoning to the proper sentence: These conclusions bring me to the heart of the reasons stated by the other members of this Court. Those reasons present five issues:

     1. The starting point issue: Whether the Court of Criminal Appeal erred in taking as its starting point, for the criticism of the sentence imposed by the sentencing judge, the quantity of the heroin the subject of the principal offence and the maximum sentences fixed by the relevant legislation for specific quantities[103];

     2. The instinctive synthesis issue: Whether the Court of Criminal Appeal erred in failing to observe a process of an "instinctive" or "intuitive" synthesis both in testing the sentence imposed by Hosking DCJ for error and in proceeding to its own substitute sentence[104];

     3. The consideration of additional offences issue: Whether the Court of Criminal Appeal erred in identifying, as part of its substitute sentence, a separate additional punishment of eighteen months to two years imprisonment for the further offences disclosed by the appellant which he asked to be taken into account in conjunction with his sentence for the principal offence[105];

     4. The manifest excess issue: Whether, whatever outcome was appropriate to the other issues in this appeal, the appellant was entitled to succeed because the re-sentencing by the Court of Criminal Appeal resulted in a sentence that was manifestly excessive and such as to attract the intervention of this Court on that ground[106]; and

     5. The proper outcome issue: Whether, if error be shown warranting intervention of this Court, the proper course is to restore the sentence imposed on the appellant by Hosking DCJ or to return the matter to the Court of Criminal Appeal[107].

107. Common ground on most issues: I agree with the conclusions stated in the joint reasons on the starting point issue and the additional offences issue. I therefore agree that the appellant's appeal must be allowed[108]. For the disposition of the appeal, I agree with the joint reasons that the proper course is to return the appeal to the Court of Criminal Appeal.

108. I am more hesitant over the manifest excess issue. It is one thing to exhibit reluctance to examine such a ground at the stage of an application for special leave to appeal to this Court. However, special leave having been granted and the matter being before this Court, manifest error (if it can be demonstrated) is a proper consideration to be taken into account. It is, after all, a well-known and unquestioned category for the appellate review of judicial discretions[109]. In my opinion, this aspect of the law of appellate reconsideration should not be excised and disregarded simply because the discretion in question concerns a judicial sentence. In the present case, the disturbance of the relativities of the sentences of the respective participants in Mr Caccamo's enterprise, and the virtual equality of the re-sentencing of the appellant and the sentence imposed on Mr Caccamo, strikes me, with respect, as manifestly erroneous. However, as nothing ultimately turns on this point, I will not press my disagreement on this issue to a dissent.

109. Nevertheless, there are important differences between the approach expressed in the joint reasons and the approach that I favour on the remaining issue. As the matter is to be returned to the Court of Criminal Appeal, the differences cannot be treated as legally immaterial. I must accordingly address these considerations in the remaining parts of these reasons.

     The "two-stage approach" versus "instinctive synthesis"

110. Origin of the controversy: The appellant complains that the Court of Criminal Appeal erred in "[f]ormulating the substituted sentence by means of a staged approach". For as long as judges have been sentencing convicted offenders for crimes, where they have had a discretion to impose a sentence within limits fixed by law, they have typically considered aggravating and mitigating circumstances when coming to their result.

111. Sometimes, judges would explain the ultimate outcome by reference to what the sentence might have been if this, or that, feature of the case had been different. There is nothing unusual in proceeding in this way. For example, in Veen v The Queen [No 2][110], decided in 1988, the joint majority reasons in this Court[111] clearly envisaged[112] such a "two-stage" or "two-tiered" approach. In a conventional fashion, the Court postulated a sentence. It then adjusted this to take into account matters special to the case. This approach reflected, and reinforced, conventional sentencing practice in Australian courts, "first determining the outer limit of the sentence and then applying mitigating factors, if any, so as to arrive at an appropriate sentence"[113].

112. In truth, this approach to sentencing did no more than to put on paper a logical process of human reasoning. Of course, the mitigating and aggravating factors would often be many and varied. But where particular considerations were clearly important, a process of reasoning would follow such as was described in Veen [No 2]. Many judges have exposed that process of reasoning in their explanations for sentence. In my view, it was honest, useful and lawful for them to do so.

113. Unfortunately, in R v Young[114], the legal waters were muddied by a decision of the Court of Criminal Appeal of Victoria. That decision was given two years after Veen [No 2]. There, that Court set its face against the stated process of reasoning in sentencing. It rejected the two-stage approach as incompatible with "long established practice in Victoria"[115]. From a practical viewpoint, it expressed its concern that such an approach was likely to result in the imposition of inadequate sentences. Revealingly, the Victorian judges also noted that by exposing the processes of reasoning of the sentencing judge in this way the approach would facilitate appellate challenges, much more difficult of success where the steps of reasoning were submerged in an outcome expressed in general terms and attributed to judicial "instinct". All appellate judges would have been aware that the more that sentencing judges exposed of the steps taken in their process of reasoning, the more likely it would be that specific error would be revealed, facilitating prisoner appeals against sentence. And sentencing appeals were viewed by many judges as a "painful" and "unrewarding" task[116].

114. The Victorian courts adhered to their opposition to the two-stage approach. Inevitably, out of necessity, trial judges in that State have bowed to the requirement to proceed in a staged way where Parliament (perhaps ignorant of the judicial minefield into which it was treading) imposed statutory obligations to reduce a sentence otherwise appropriate for a plea of guilty[117]. Such cases apart, the Victorian judges continued to reject the two-stage approach[118]. Even in a case where statute[119] appeared to require identification of a sentence and adjustment for the statute's purposes, the Victorian Court of Criminal Appeal insisted on the adoption of what it called an "instinctive synthesis" of all relevant matters, including such adjustments.

115. Developments in State jurisdictions: This Court noticed this controversy in Bugmy v The Queen[120], a Victorian appeal. It did not resolve it there. Meantime, the controversy simmered in other Australian States. Most of the judges of other States who passed upon the issue expressed themselves unconvinced by the Victorian approach in Young.

116. Thus, the Court of Criminal Appeal of the Northern Territory in R v Raggett[121] rejected the criticism about adopting the two-stage approach. Likewise, the strict embargo on the two-tiered approach was not followed in New South Wales in R v Gallagher[122]. In South Australia, the two-tiered approach was well established in the practice of judicial sentencing. This was especially so for discounts for pleas of guilty and for assistance to authorities. King CJ, who knew a great deal about criminal law and practice, endorsed the two-stage approach in R v Shannon[123]. It was also reflected in his Honour's highly influential reasons in R v Osenkowski[124], where he defended the entitlement of sentencing judges "occasionally to correct a sentence"[125] out of a sense of reasoned leniency in the particular circumstances.

117. The approach in Young fared no better in Western Australia[126], although it gathered some support in the Tasmanian Court of Criminal Appeal in Pavlic v The Queen[127]. This was not, however, without a strong dissent on this point by Slicer J. By the late 1990s, no other court of criminal appeal of this nation had clearly embraced the anathema in Young on the two-stage approach. To the contrary, many judges, highly experienced in sentencing at trial and on appeal, rejected that approach. For identified reasons of "social utility and public policy"[128], they saw value in exposing the process of reasoning towards their sentences and (whilst not obliging such a course as an absolute rule) they saw utility in identifying specifically, in quantitative or percentage terms, discounts for various considerations such as pleas of guilty and specific assistance to the authorities[129]. By 1999, so far as the rest of Australia was concerned, Young looked dead in the Yarra River water.

118. "Wrong in principle?": It was at this stage, in 1999, that Hayne J, in AB v The Queen[130], in dissenting reasons in this Court, indicated his adherence to an approach similar to that expressed in Young and to the earlier statement in the Victorian Full Court in R v Williscroft[131]. There that Court had said that it is "profitless ... to attempt to allot to the various considerations their proper part in the assessment of the particular punishments". Instead, according to Williscroft, the sentence to be imposed "represents the sentencing judge's instinctive synthesis" of relevant considerations.

119. In his reasons in AB, McHugh J, also in dissent in that case, endorsed a similar approach, holding that the two-stage approach was "plainly unsuited to the sentencing process"[132]. There this minor judicial controversy might have rested, but for the joint reasons of Gaudron, Gummow and Hayne JJ in Wong v The Queen[133]. An extract from those reasons appears in the joint reasons in this case[134]. The two-stage approach was there castigated as "wrong in principle"[135], in terms directly traceable to the idiosyncratic view expressed in Young, which in turn built on the "instinctive" approach endorsed earlier in Victoria in Williscroft. Although in Wong it is stated that the intermediate appellate courts of Australia were, by that time, "clearly against adopting two-stage sentencing and favour[] the instinctive synthesis approach"[136], I have endeavoured to show (and many more cases could be added to my list) that this was not a correct representation of the state of decisional authority.

120. Now, building upon this highly unstable foundation of judicial reasoning, the dictum in Wong at last gathers up a majority of this Court, for it is apparently endorsed in the joint reasons and the reasons of McHugh J. Because I do not agree with it either as a matter of analysis of Australian judicial authority or as a matter of legal principle and policy, I must voice my contrary opinion. The fact that, in the joint reasons in this case, there is a substantial retreat from the strict anathema expressed in Young does not mean that the error of the joint reasons in Wong, now repeated with added qualifiers, should pass unremarked[137].

121. Decisions since Wong: It was inevitable, following the differing opinions expressed in this Court in Wong, that intermediate appellate courts in Australia, reviewing sentences, should struggle to accommodate the differing views stated (although without binding authority) by judges of this Court. Naturally, none of the courts below wished needlessly to expose themselves to the peril of reversal should the approach in Young ultimately prevail in this Court. Yet none (so far as my reading shows) was willing to accept that a two-stage approach was universally impermissible ("wrong in principle") or that the complexities of sentencing could be adequately hidden by adopting a judicial formula such as the so-called "instinctive synthesis".

122. Specifically, in many decisions, the intermediate courts saw nothing wrong (and much that was advantageous) in the explicit identification of the precise discount to be allowed, in particular cases, for pleas of guilty and for assistance to the authorities. They therefore did what was sensible in the circumstances. They adhered, in fact, to the two-stage approach in those and other instances of sentencing. However, they accepted that this was not a universal approach of sentencing but one specific to the ad hoc instances where it was appropriate or at least permissible. The formula "two-stage approach" was sometimes replaced by descriptions such as "sequential process"[138]. The judges occasionally confessed (as Hulme J did in this case) that they found it "difficult to understand" how the "instinctive synthesis" approach could be applied, or how it could result in a "single appropriate sentence" unless an hypothesised starting point were taken "against which the factors of assistance and of the plea could be considered"[139]. For the specification of particular discounts (and hence the necessity of a kind of two-stage approach) a unanimous Court of Criminal Appeal of South Australia in R v Place[140] clearly remained unconvinced by the approach demanded in Young. Indeed, it was dismissive of the favour it had, by that time, gathered in this Court, then still short of a majority. In Place, the South Australian Court of Criminal Appeal said[141]:

     "For these reasons, in our opinion the current practice should continue and this Court should continue to encourage sentencing courts to identify the specific reduction given in respect of a plea of guilty."

123. As a matter of principle, the same approach applied in South Australia for the consideration of assistance to authorities mentioned earlier in the reasons in Place[142]. But once that position was reached, as a matter of logic and principle, the same approach would necessarily apply to any other distinct factor in sentencing, important to the particular case, that caused a measurable and clearly identifiable adjustment to the sentence that warranted explicit mention in discharging the sentencing function according to law, not in accordance with supposed judicial "instinct".

124. Where so many judges in Australia, experienced in criminal trials and in sentencing, have expressed their disagreement with the approaches derived from Williscroft and Young, it is undesirable, in my respectful opinion, for this Court (even in the present watered-down version) to impose those authorities on sentencing judges throughout the Commonwealth.

125. Inconsistency with statutory transparency: An additional reason, which should cause hesitation on our part in this respect, is the growing move of federal and State legislatures in Australia to spell out specific considerations that are to be taken into account in judicial sentencing. This is obviously the purpose of s 21A of the Sentencing Act which was overlooked in this case. But those provisions are simply examples of a multitude of contemporary statutory requirements, in virtually every Australian jurisdiction, federal, State and Territories, obliging sentencing courts and courts of criminal appeal to pay regard to aggravating and mitigating factors. Sometimes, these will suggest the need for adjustment stated in quantitative or percentage terms. Always, they postulate the contemplation of an hypothesised norm that is adjusted up and down. Indeed, in some instances, such adjustments are expressly required by the legislation, such as is the case under the Sentencing Act[143].

126. In particular circumstances, the introduction of standard non-parole sentencing obliges judges, in effect, to adopt a two-stage approach[144]. In such cases, it is impossible to conceive that a purely instinctive synthesis or "single-tiered approach" could be taken to the sentencing of a prisoner affected by the statute. The sentencing judge is effectively obliged to identify the "standard non-parole period"[145]. It then becomes the statutory reference point relating to the "middle of the range of" objective seriousness of the offence. The sentencing judge is then required to take into account the other matters referred to in the Sentencing Act, in order to arrive at an appropriate non-parole period and hence the resulting balance of the sentence[146]. A clearer example of a two-stage approach could not be imagined.

127. In this statutory environment, given the first duty of sentencing judges to conform to applicable parliamentary law, the instances for a single-tiered and purely instinctive synthesis in sentencing will now be increasingly rare, if ever they existed. If legislation obliges identification of adjustments to "standard ... period" sentences, it is questionable that the common law of Australia should now, belatedly, embrace any different rule where it has not previously been regarded by judges as a universal obligation.

128. Considerations of function and principle: Whilst I recognise that the modified version of the prohibition against two-tiered sentencing, now adopted in the joint reasons in this appeal, permits exceptions and acknowledges, in effect, that instances will exist where "some indulgence in an arithmetical process" will pass muster[147], the continued endorsement of the discredited view of sentencing as an "instinctive synthesis" remains to undermine this ultimate acknowledgment of the inescapable reality.

129. With all respect to those of the different opinion, the phrase "instinctive synthesis" sends quite the wrong signals for the law of sentencing in Australia. Who are those who have the "instincts" in question? Only the judges. This is therefore a formula that risks endorsement of the deployment of purely personal legal power. It runs contrary to the tendency in other areas of the law, notably administrative law, to expose to subsequent scrutiny the use of public power by public officials[148]. It is contrary to the insistence of Australian courts[149], including this Court[150], that judicial officers must give reasons for their decisions. At this stage in the development of the Australian law of sentencing, this Court should be encouraging, not impeding, transparency and accountability of judicial decision-making[151]. I remain of the view that "[i]t is too late (and undesirable) to return to unexplained judicial intuition"[152]. Talk of "instinctive synthesis" is like the breath of a bygone legal age. It resonates with a claim, effectively, to unexplainable and unreviewable power.

130. It is for these reasons that the supposed "instinctive synthesis", as an explanation of the judicial task in sentencing, has been criticised by knowledgeable experts in criminal law and sentencing[153]. All of those experts know, and recognise, that there are limits to the explanation of reasons for a given sentence. Ultimately, unless the law itself fixes the sentence, judgment is invoked. However, as the present appeal demonstrates, appellate courts expounding general principles should encourage revelation at least of the important adjustments that are made by a sentencing judge. They should not be encouraging the thought that there descends upon a judicial officer, following appointment, a mystical "instinct" or "intuition" that ensures that he or she will get the sentence right "instinctively". That approach discourages explanation of the logical and rational process that led to the sentence, so far as it can reasonably be given and is useful.

131. Functional analysis also suggests that talk of judicial "instinct" is ill-advised. If, in reasoning, the judicial officer does make a significant adjustment for a particular factor - measurable in the judge's opinion in quantitative or percentage terms - the choice before the law is whether that factor should be specifically exposed in the reasons or not. There are many grounds of policy and principle, in such circumstances, why it should be[154]. If it is not identified, the risk that arises is that identified by Hulme J in the Court of Criminal Appeal in this case. Some judges will feel that it is safer, wiser or even essential to keep the process of reasoning secret. That course is good neither for the parties, nor for the community, nor for the discharge of the functions of sentencing, nor for appellate review[155]. With some judicial officers, talk of "instinct" and pure "intuition" might be understood as endorsing a process of sentencing that involves little more than plucking a figure from the air, to use Hulme J's telling expression[156]. Such an arbitrary exercise of public power is to be discouraged, not endorsed by the use by this Court of phrases such as "instinctive synthesis".

132. Semantics and substance: I have previously suggested that some of the debates over the two-stage approach and instinctive or intuitive synthesis may be semantic, not substantive[157]. That remains my view. To this extent, I agree with what is said in that part of the joint reasons[158]. But a sticking point remains, for I cannot accept a Williscroft "instinct" or a Young prohibition on two-stage reasoning as sentencing principles, where a more transparent course is available, appropriate and more conformable with modern legal principles governing the deployment of public power. To say the least, there have been important developments in the subjection of uncontrolled discretions to judicial analysis since R v Geddes[159] was written[160]. Fundamentally, such developments derive from a principle that lies at the heart of the Australian Constitution and its system of democratic and accountable government. Intuitive and instinctive power is not now in favour. The rule of law stands in its place[161].

133. I agree that there is no single correct sentence (unless it is lawfully fixed by Parliament). I also agree that sentencing is not a mechanical, numerical, arithmetical or rigid activity in which one starts from the maximum fixed by Parliament and works down in mathematical steps[162]. The process is not so scientific. Because there are a multitude of factors to be taken into account, many of them pulling successively in opposite directions, the evaluation, in terms of time of imprisonment, quantity of fine or other sanction, is necessarily imprecise[163]. Human judgment is inevitably invoked. In sentencing there is sometimes a legitimate role for differences of judicial view. These may occasionally favour the extension of leniency, as Osenkowski[164] shows. Necessarily, there must also be room for the views of a judicial officer who takes a more punitive view of all of the relevant considerations in the case. So long as all relevant considerations are given due attention, the discretionary character of sentencing will inhibit appellate interference.

134. This said, there are outer boundaries. They control the scope for judicial officers to indulge individual idiosyncrasies. Sentencing appeals afford a protection against miscarriages of justice that weigh heavily on the liberty of the individual affected. The appellate court should be attentive to the possibility of error. But it is not, in my view, an error of sentencing principle for the sentencing judge to proceed in two or more stages. Exposure of particular discounts - for a plea of guilty, the provision of assistance to authorities or other considerations that seem most significant - is not compulsory unless statute makes it so. But neither does it constitute an error of sentencing principle as such.

135. Judicial officers engaged in sentencing should be encouraged to reveal their processes of reasoning. Simply to assert that they have considered a list of relevant matters, without identifying, in general terms, the weight that has been given to the most important of them, may represent an error in sentencing. The generalised assertion by the sentencer that he or she has acted on "instinct", "intuition" or personal experience or the experience in the courts, is not now enough, in my opinion, to meet the standards of reasoning in sentencing that we have come to expect in Australia. Honesty and transparency in the provision of reasons is the hallmark of modern judicial administration. Not judicial "instinct".

     Conclusion: a needless diversion

136. Limited residue of the prohibition: Where, then, have we arrived at the end of this judicial journey? The joint reasons continue to chastise the "two-tiered approach"[165]. Yet if it is merely a "sequential" approach, involving distinct factors, it is apparently unobjectionable. The difference will usually be illusory. Moreover, there are now "[n]o universal rules"[166]. This is at least an advance on Young and the earlier unyielding prohibition upon a staged explanation of the ultimate sentence imposed by a judge. However, there is still talk of "instinctive synthesis"[167]. Yet this too must apparently be reconciled with the obligation of public decision-makers to transparency and also with the specific judicial duty to provide proper reasons[168].

137. In the end, even the postulated process of "instinctive synthesis" to the judicial outcome is not apparently to be confused with "mysterious" and "arcane" activities limited to judges[169]. Perhaps, in the end, the "instinctive synthesis" means nothing more than that the sentencing judge is to take everything relevant into account and to reach a final judgment. But this is what judges have always had to do. So what does the reference to "instinctive" add, except to distract?

138. All that seems to be left from the original imperatives, traced to the decisions in Williscroft and Young, is a prohibition on mathematical adjustment in deriving the ultimate sentence imposed on an offender. Yet even this is not now absolute. Specification, in a staged or sequential approach, of the degree of reduction of what would otherwise have been the penalty for a plea of guilty is, it seems, sometimes permissible[170]. So presumably is re-adjustment for any assistance to authorities. So indeed, by statutes in many parts of Australia, must now be specific reductions and adjustments expressed in terms of identified quantification or percentages. Even occasionally (albeit in unexplained circumstances) arithmetical indulgence will now, it seems, be overlooked. However, preferably that will happen only where the factors adjusted are comparatively few and the case is "simple"[171].

139. So analysed, the residue of this judicial debate over twenty years - in this Court over the past five years - is revealed for what it is. Australian judges must now express their obeisance to an "instinctive synthesis" as the explanation of their sentencing outcomes. It might be prudent for them to avoid mention of "two stages" or of mathematics. Yet in many instances (and increasingly by statutory prescription) if judges do so, no error of sentencing principle will have occurred. Such mention may, in fact, sometimes even be required[172]. The lofty and absolute prescriptions of Williscroft and Young remain in place like the two vast and trunkless legs of stone of Ozymandias[173]. But, with all respect, they are now beginning to look just as lifeless. One day, I expect that travellers to the antique land of this part of the law of sentencing will walk this way without knowing that the two proscriptions once were there.

140. The ironic disclosure of error: By virtue of the transparent approach taken correctly, in my view, by the Court of Criminal Appeal, it is apparent that that Court erred in adopting the wrong starting point for consideration of the appellant's sentence for the principal offence. That error warrants correction. For this reason, the appeal must succeed and the matter must be remitted to the Court of Criminal Appeal for disposition of the appeal in light of these reasons.

141. However, the appellant's specific complaint that the Court of Criminal Appeal erred in adopting the "staged approach" in my view fails. But for that approach, the appellant's appeal to this Court would probably never have been heard, and the errors that now occasion the appellant's success would not have been revealed.

     Orders

142. For the foregoing reasons, I agree in the orders proposed in the joint reasons.

[1] Inserted by the Crimes (Sentencing Procedure) Amendment (General Sentencing Principles) Act 2002 (NSW), Sched 1, Item [1]. Repealed and substituted, with effect from 1 February 2003 by the Crimes (Sentencing Procedure) Amendment (Standard Minimum Sentencing) Act 2002 (NSW), Sched 1, Item [2].

[2] R v Markarian [2003] NSWCCA 8; (2003) 137 A Crim R 497.

[3] [2003] NSWCCA 8; (2003) 137 A Crim R 497 at 501 [17].

[4] [2003] NSWCCA 8; (2003) 137 A Crim R 497 at 501 [18].

[5] See Wong v The Queen [2001] HCA 64; (2001) 207 CLR 584.

[6] R v Rushby [1977] 1 NSWLR 594; Dodd (1991) 57 A Crim R 349; Bimahendali [1999] NSWCCA 409; (1999) 109 A Crim R 355; R v Whyte [2002] NSWCCA 343; (2002) 55 NSWLR 252.

[7] [2003] NSWCCA 8; (2003) 137 A Crim R 497 at 502-503 [23].

[8] [2003] NSWCCA 8; (2003) 137 A Crim R 497 at 503-504 [24]- [28].

[9] Veen v The Queen [No 2] [1988] HCA 14; (1988) 164 CLR 465 at 477.

[10] Perese (2001) 126 A Crim R 508; Kay (2002) 132 A Crim R 72; R v AEM [2002] NSWCCA 58.

[11] [2003] NSWCCA 8; (2003) 137 A Crim R 497 at 505 [33]- [35].

[12] See R v Whyte [2002] NSWCCA 343; (2002) 55 NSWLR 252 and the cases therein cited.

[13] [2003] NSWCCA 8; (2003) 137 A Crim R 497 at 505-506 [37]- [38].

[14] [1988] HCA 14; (1988) 164 CLR 465.

[15] [1971] 1 NSWLR 247.

[16] [1991] 1 VR 717 at 722.

[17] [2003] NSWCCA 8; (2003) 137 A Crim R 497 at 506 [40].

[18] R v Le Cerf (1975) 13 SASR 237.

[19] [1999] NSWCCA 3 at [33].

[20] [2003] NSWCCA 8; (2003) 137 A Crim R 497 at 507 [45].

[21] [2003] NSWCCA 8; (2003) 137 A Crim R 497 at 509 [55]- [56].

[22] [2001] HCA 64; (2001) 207 CLR 584 at 609 [70].

[23] [1936] HCA 40; (1936) 55 CLR 499 at 504-505.

[24] Pearce v The Queen (1998) 194 CLR 610 at 624 [46].

[25] Johnson v The Queen [2004] HCA 15; (2004) 78 ALJR 616 at 618 [5] per Gleeson CJ, 624 [26] per Gummow, Callinan and Heydon JJ; [2004] HCA 15; 205 ALR 346 at 348, 356.

[26] [1999] HCA 29; (1999) 195 CLR 665 at 671-672 [15].

[27] Stockdale and Devlin, Sentencing, (1987), pars 1.16-1.18.

[28] The maximum selected by his Honour was not, as will appear, the maximum available in respect of the principal offence.

[29] [2003] NSWCCA 8; (2003) 137 A Crim R 497 at 506 [37].

[30] [2001] HCA 64; (2001) 207 CLR 584 at 611-612 [74]- [76] (some footnotes omitted).

[31] (2000) 49 NSWLR 383.

[32] [1999] HCA 46; (1999) 198 CLR 111.

[33] (1991) 23 NSWLR 220.

[34] [2002] NSWCCA 142; (2002) 54 NSWLR 300.

[35] [2002] NSWCCA 343; (2002) 55 NSWLR 252.

[36] [2002] SASC 101; (2002) 81 SASR 395.

[37] [2002] SASC 101; (2002) 81 SASR 395 at 424-425 [80]- [83].

[38] R v Markarian [2003] NSWCCA 8; (2003) 137 A Crim R 497 at 505 [33].

[39] The Commonwealth v Bank of NSW [1949] HCA 47; (1949) 79 CLR 497 at 642; [1950] AC 235 at 313.

[40] [1999] HCA 46; (1999) 198 CLR 111 at 120-123 [13]- [19].

[41] "The Weighing of Evidence and the Determinants of Confidence", (1992) 24 Cognitive Psychology 411 at 413.

[42] [1990] VicRp 84; [1990] VR 951 at 960.

[43] This was the name that Gauss used in signing his works although he was christened Johann Friedrich Carl Gauss. Gauss made many mathematical discoveries. In the realm of statistics, he invented the "least squares" method that is indispensable "in all work where the 'most probable' value of anything that is measured is to be inferred from a large number of measurements." Another statistical tool that we owe to Gauss is "[t]he Gaussian law of normal distribution of errors and its accompanying bell-shaped curve [that] is familiar today to all who handle statistics" (Bell, Men of Mathematics, (1937) at 227).

[44] [2001] HCA 64; (2001) 207 CLR 584 at 611 [74].

[45] [2003] NSWCCA 8; (2003) 137 A Crim R 497 at 506 [38].

[46] [2003] NSWCCA 8; (2003) 137 A Crim R 497 at 506 [39].

[47] [2003] NSWCCA 8; (2003) 137 A Crim R 497 at 506 [40].

[48] R v Le Cerf (1975) 13 SASR 237.

[49] [2003] NSWCCA 8; (2003) 137 A Crim R 497 at 505 [32].

[50] [1988] HCA 14; (1988) 164 CLR 465.

[51] [1971] 1 NSWLR 247.

[52] [2003] NSWCCA 8; (2003) 137 A Crim R 497 at 506 [38].

[53] [2003] NSWCCA 8; (2003) 137 A Crim R 497 at 506 [39].

[54] [2003] NSWCCA 8; (2003) 137 A Crim R 497 at 506 [37].

[55] [2003] NSWCCA 8; (2003) 137 A Crim R 497 at 507 [44].

[56] [2003] NSWCCA 8; (2003) 137 A Crim R 497 at 507 [45].

[57] [2003] NSWCCA 8; (2003) 137 A Crim R 497 at 507 [46].

[58] [2003] NSWCCA 8; (2003) 137 A Crim R 497 at 508 [47].

[59] [2003] NSWCCA 8; (2003) 137 A Crim R 497 at 508 [48].

[60] [2003] NSWCCA 8; (2003) 137 A Crim R 497 at 508 [49].

[61] [2003] NSWCCA 8; (2003) 137 A Crim R 497 at 509 [55].

[62] (1936) 36 SR (NSW) 554 at 555-556.

[63] [1975] VicRp 27; [1975] VR 292 at 300.

[64] (1936) 36 SR (NSW) 554.

[65] R v Goodrich (1952) 70 WN (NSW) 42.

[66] R v Cooke (1955) 72 WN (NSW) 132.

[67] R v Herring (1956) 73 WN (NSW) 203.

[68] (1956) 73 WN (NSW) 203 at 205.

[69] (1956) 73 WN (NSW) 203 at 205.

[70] [1988] HCA 14; (1988) 164 CLR 465.

[71] [1990] VicRp 84; [1990] VR 951 at 957.

[72] [1990] VicRp 84; [1990] VR 951 at 957.

[73] Veen v The Queen [1979] HCA 7; (1979) 143 CLR 458.

[74] The Supreme Court recently considered the constitutionality of these Guidelines in United States v Booker 73 USLW 4056 (2005).

[75] [1988] HCA 14; (1988) 164 CLR 465 at 476.

[76] Cameron v The Queen (2002) 209 CLR 339 at 345-346 [22] per Gaudron, Gummow and Callinan JJ.

[77] [2003] NSWCCA 8; (2003) 137 A Crim R 497 at 509 [55].

[78] [1988] HCA 14; (1988) 164 CLR 465 at 472.

[79] [1936] HCA 40; (1936) 55 CLR 499 at 505 (referred to in the joint judgment at [25]).

[80] R v Markarian [2003] NSWCCA 8; (2003) 137 A Crim R 497.

[81] As to specific error and manifest excess, see AB v The Queen [1999] HCA 46; (1999) 198 CLR 111 at 160 [130] per Hayne J.

[82] Joint reasons at [2]-[19].

[83] [2004] HCATrans 329 at 1936.

[84] See joint reasons at [2], [11] referring to Markarian [2003] NSWCCA 8; (2003) 137 A Crim R 497 at 501 [17].

[85] See joint reasons at [3].

[86] Lowndes v The Queen [1999] HCA 29; (1999) 195 CLR 665 at 671-672 [15]. See also Dinsdale v The Queen [2000] HCA 54; (2000) 202 CLR 321 at 339-340 [57]- [58], 340-341 [62]; cf AMS v AIF (1999) 199 CLR 160 at 222 [183]. The principle was stated earlier in Norbis v Norbis [1986] HCA 17; (1986) 161 CLR 513 at 517-519; cf Wong v The Queen (2001) [2001] HCA 64; 207 CLR 584 at 622-623 [104].

[87] Conway v The Queen (2002) 209 CLR 203 at 227 [66]; cf Visy Paper Pty Ltd v Australian Competition and Consumer Commission [2003] HCA 59; (2003) 216 CLR 1 at 10 [24], 25 [74] and cases there cited.

[88] Joint reasons at [5].

[89] Joint reasons at [10]-[19].

[90] eg AB [1999] HCA 46; (1999) 198 CLR 111 at 150 [102]; Wong [2001] HCA 64; (2001) 207 CLR 584 at 622 [102]- [103]; Cameron v The Queen (2002) 209 CLR 339 at 362 [70]-[71] and Johnson v The Queen [2004] HCA 15; (2004) 78 ALJR 616 at 626 [40]; [2004] HCA 15; 205 ALR 346 at 358-359.

[91] Markarian [2003] NSWCCA 8; (2003) 137 A Crim R 497 at 505 [33]- [35], cited in the joint reasons at [14].

[92] Markarian [2003] NSWCCA 8; (2003) 137 A Crim R 497 at 505 [33].

[93] It is very rare for the High Court to give relief for "manifest [as distinct from specific] error" of sentencing: AB [1999] HCA 46; (1999) 198 CLR 111 at 126 [30]; see also Postiglione v The Queen [1997] HCA 26; (1997) 189 CLR 295 at 337.

[94] Joint reasons at [44].

[95] Joint reasons at [29].

[96] [1999] HCA 29; (1999) 195 CLR 665 at 671-672 [15].

[97] AMS (1999) 199 CLR 160 at 222 [183]; cf Coal and Allied Operations Pty Ltd v Australian Industrial Relations Commission [2000] HCA 47; (2000) 203 CLR 194 at 225-226 [77]; The Commonwealth v Yarmirr [2001] HCA 56; (2001) 208 CLR 1 at 119 [266].

[98] The Sentencing Act, s 21A(4), as then enacted, stated: "The matters to be taken into account by a court under this section are in addition to any other matters that are required or permitted to be taken into account by the court under this Act or any other law." This requirement has now become part of s 21A(1); cf R v Way [2004] NSWCCA 131; (2004) 60 NSWLR 168 at 189 [103].

[99] Joint reasons at [3].

[100] House v The King [1936] HCA 40; (1936) 55 CLR 499 at 505 per Dixon, Evatt and McTiernan JJ: "[I]f [the judge] does not take into account some material consideration, then his determination should be reviewed and the appellate court may exercise its own discretion in substitution for his if it has the materials for doing so."

[101] Joint reasons at [3].

[102] Section 101A of the Sentencing Act (which commenced after the hearing of the Crown's appeal but before the handing down of the Court of Criminal Appeal's decision) provides that the effect of failure to comply with the Act may be considered by an appellate court in any appeal against sentence even if the Act declares that the failure to comply does not invalidate the sentence. Certain sections so provide. See, for example, s 44(3) of the Sentencing Act, providing that the failure of a court to set an appropriate term for the balance of the sentence with respect to the non-parole period does not invalidate a sentence.

[103] Joint reasons at [20], [30]-[34]; reasons of McHugh J at [50].

[104] Joint reasons at [35]-[39]; reasons of McHugh J at [50].

[105] Joint reasons at [40]-[43].

[106] Joint reasons at [44].

[107] Joint reasons at [46].

[108] Joint reasons at [45]; reasons of McHugh J at [85].

[109] House v The King [1936] HCA 40; (1936) 55 CLR 499 at 505.

[110] [1988] HCA 14; (1988) 164 CLR 465.

[111] Per Mason CJ, Brennan, Dawson and Toohey JJ.

[112] [1988] HCA 14; (1988) 164 CLR 465 at 476-478. See also Baumer v The Queen [1988] HCA 67; (1988) 166 CLR 51 at 56-58.

[113] Bugmy v The Queen [1990] HCA 18; (1990) 169 CLR 525 at 535.

[114] [1990] VicRp 84; [1990] VR 951.

[115] [1990] VicRp 84; [1990] VR 951 at 960-961.

[116] Lord Kilbrandon, "Children in Trouble", (1966) 6 British Journal of Criminology 112 at 122; Kirby, "Sentencing Reform: Help in the 'Most Painful' and 'Unrewarding' of Judicial Tasks", (1980) 54 Australian Law Journal 732 at 734.

[117] See, for example, Penalties and Sentences Act 1985 (Vic), s 4(2) (repealed); Tierney (1990) 51 A Crim R 446; cf Sentencing Act 1991 (Vic), ss 5(2AB), 5(2AC), 5(2)(e).

[118] See, for example, R v Nagy [1992] 1 VR 637. The history is traced in Punch v The Queen (1993) 9 WAR 486 at 493-496 per Murray J.

[119] Crimes Act 1914 (Cth), ss 16A and 21E. See Nagy [1992] 1 VR 637.

[120] [1990] HCA 18; (1990) 169 CLR 525 at 535-536.

[121] (1990) 101 FLR 323 at 334-335; 50 A Crim R 41 at 51-52. See also R v Mulholland (1991) 1 NTLR 1 at 14-15.

[122] (1991) 23 NSWLR 220 at 230 per Gleeson CJ.

[123] (1979) 21 SASR 442 at 452-453.

[124] (1982) 30 SASR 212.

[125] (1982) 30 SASR 212 at 212-213.

[126] McKenna v The Queen (1992) 7 WAR 455 at 467-468; Punch (1993) 9 WAR 486 at 493-496, 503; Verschuren v The Queen (1996) 17 WAR 467 at 470-474, 480-491.

[127] [1995] TASSC 96; (1995) 5 Tas R 186.

[128] Pavlic [1995] TASSC 96; (1995) 5 Tas R 186 at 206 per Slicer J.

[129] Thus in Verschuren (1996) 17 WAR 467 at 473, Malcolm CJ expressed agreement with the reasons of Slicer J in Pavlic. See also R v Place [2002] SASC 101; (2002) 81 SASR 395 at 416 [55]- [56].

[130] [1999] HCA 46; (1999) 198 CLR 111 at 156-157 [115]- [120].

[131] [1975] VicRp 27; [1975] VR 292 at 300; cf Place [2002] SASC 101; (2002) 81 SASR 395 at 413-414 [47]- [48].

[132] [1999] HCA 46; (1999) 198 CLR 111 at 121 [16].

[133] [2001] HCA 64; (2001) 207 CLR 584 at 611 [75]- [77].

[134] Joint reasons at [37].

[135] [2001] HCA 64; (2001) 207 CLR 584 at 612 [76].

[136] Wong [2001] HCA 64; (2001) 207 CLR 584 at 611 [76].

[137] See Cameron (2002) 209 CLR 339 at 362 [70]-[71]; Johnson [2004] HCA 15; (2004) 78 ALJR 616 at 626-627 [40]- [42]; [2004] HCA 15; 205 ALR 346 at 358-359.

[138] See R v Garforth unreported, New South Wales Court of Criminal Appeal, 23 May 1994 at 6 per Gleeson CJ, McInerney and Mathews JJ; Way [2004] NSWCCA 131; (2004) 60 NSWLR 168 at 190 [112]; cf R v Thomson (2000) 49 NSWLR 383 at 396 [57]; R v Sharma [2002] NSWCCA 142; (2002) 54 NSWLR 300 at 305 [24], 307 [31].

[139] MacDonnell [2002] NSWCCA 34; (2002) 128 A Crim R 44 at 54 [59] per Wood CJ at CL for the Court. See also Howie, "Criminal Law Update 2004", (2004) 7 Judicial Review 89 at 103-104.

[140] [2002] SASC 101; (2002) 81 SASR 395.

[141] [2002] SASC 101; (2002) 81 SASR 395 at 425 [83] per Doyle CJ, Prior, Lander and Martin JJ (Gray J concurring).

[142] [2002] SASC 101; (2002) 81 SASR 395 at 417 [59].

[143] For example, Sentencing Act, s 21A (requiring certain aggravating and mitigating factors to be taken into account in determining sentence); Sentencing Act, s 22 (a court may take into account a guilty plea and in doing so impose a "lesser penalty"); Sentencing Act, s 23 (assistance to authorities); and Sentencing Act, s 33 (taking into account a further offence which may lead to a longer sentence).

[144] Crimes (Sentencing Procedure) Amendment (Standard Minimum Sentencing) Act 2002 (NSW) amending the Sentencing Act to insert ss 54A-54D in the Act.

[145] Sentencing Act, s 54A.

[146] Sentencing Act, ss 21A, 44, discussed in Way [2004] NSWCCA 131; (2004) 60 NSWLR 168; see also Traynor and Potas, "Sentencing Methodology: Two-tiered or Instinctive Synthesis?", (2002) 25 Sentencing Trends and Issues (Judicial Commission of New South Wales) 1 at 14.

[147] Joint reasons at [39].

[148] Davis, Discretionary Justice: A Preliminary Inquiry, (1971) at 31; Dworkin, Taking Rights Seriously, (1978) at 31-33; Galligan, Discretionary Powers: A Legal Study of Official Discretion, (1986) at 17-22; cf Osmond v Public Service Board of New South Wales [1984] 3 NSWLR 447 at 462-464 citing Wade, Administrative Law, 5th ed (1982) at 486.

[149] Pettitt v Dunkley [1971] 1 NSWLR 376 at 387-388 per Moffitt JA.

[150] Public Service Board of NSW v Osmond (1986) 159 CLR 656 at 666-667.

[151] Pearce v The Queen (1998) 194 CLR 610 at 624 [46]; Wong [2001] HCA 64; (2001) 207 CLR 584 at 621-622 [101]- [102]; Weininger v The Queen [2003] HCA 14; (2003) 212 CLR 629 at 652 [75].

[152] AB [1999] HCA 46; (1999) 198 CLR 111 at 150 [102].

[153] eg Leader-Elliott, "Instinctive synthesisers in the High Court", (2002) 26 Criminal Law Journal 5; Bagaric, "Sentencing: The Road to Nowhere", [1999] SydLawRw 23; (1999) 21 Sydney Law Review 597; Bagaric and Edney, "What's instinct got to do with it? A blueprint for a coherent approach to punishing criminals", (2003) 27 Criminal Law Journal 119; cf Fox and Freiberg, Sentencing: State and Federal Law in Victoria, 2nd ed (1999) at 195-196 [3.302], 202 [3.307]. The objection to the two-stage approach is also inconsistent with approaches of final courts overseas. See eg R v McDonnell [1997] 1 SCR 948 at 986-989 [57]-[61].

[154] Cameron (2002) 209 CLR 339 at 362 [70].

[155] See above these reasons at [96]-[98].

[156] Markarian [2003] NSWCCA 8; (2003) 137 A Crim R 497 at 505 [33].

[157] Cameron (2002) 209 CLR 339 at 362 [71]; see also Punch (1993) 9 WAR 486 at 494.

[158] Joint reasons at [36].

[159] (1936) 36 SR (NSW) 554 at 555-556.

[160] See reasons of McHugh J at [65].

[161] Australian Communist Party v The Commonwealth [1951] HCA 5; (1951) 83 CLR 1 at 193; Kartinyeri v The Commonwealth [1998] HCA 22; (1998) 195 CLR 337 at 381 [89]; Plaintiff S157/2002 v The Commonwealth (2003) 211 CLR 476 at 513-514 [103]-[104] per Gaudron, McHugh, Gummow, Kirby and Hayne JJ.

[162] AB [1999] HCA 46; (1999) 198 CLR 111 at 121-122 [16]; Wong [2001] HCA 64; (2001) 207 CLR 584 at 611 [74]- [75], 612 [77]; R v Whyte [2002] NSWCCA 343; (2002) 55 NSWLR 252 at 278 [160]- [166].

[163] Weininger [2003] HCA 14; (2003) 212 CLR 629 at 645 [50].

[164] (1982) 30 SASR 212 at 212-213 per King CJ.

[165] Joint reasons at [37].

[166] Joint reasons at [36].

[167] Joint reasons at [35]-[39].

[168] Joint reasons at [39].

[169] See joint reasons at [39].

[170] Joint reasons at [38].

[171] Joint reasons at [39].

[172] When prescribed by statute.

[173] Shelley, "Ozymandias", reproduced in The Norton Anthology of English Literature, 6th ed (1993), vol 2 at 672.