Judicial implementation of legislative policy: Ruminations on the impact of the Sentencing Act 2002 on sentencing practice and prison musters

From the Wellington District Law Society Seminar
3rd April 2006

Introduction

  1. Those who promote legislation which will affect the conduct and outcomes of everyday litigation necessarily rely heavily on the willingness and ability of the courts to implement the underlying legislative policies. So judicial implementation of legislative policy is an important topic.
  1. I have chosen the Sentencing Act 2002 (to which I will generally refer as "the Act") as the focus of my discussion. In part, this is because I am reasonably familiar with its operation. As well, the issues which arise in relation to the Act are reasonably straight-forward and thus accessible and intelligible to a general legal audience. They also have a contemporary importance in light of concerns over our current rates of incarceration.
  1. I stress at the outset that implementation means more than just interpretation.
  1. There are some issues of interpretation which have been thrown up by the Sentencing Act, for instance as to the meaning of s 30(1). This subsection re-enacted what was s 13A of the Criminal Justice Act 1985 which, with some limitations, prohibited the imprisonment of defendants who had not been legally represented when at risk of conviction. This section was given an interpretation by the Court of Appeal in Parkhill v Ministry of Transport[2] which was not particularly literal or rights-centred but nonetheless appeared to reflect the underlying parliamentary intention. In R v Condon,[3] the Court of Appeal applied the Parkhill approach to the new section. The Supreme Court has now granted leave to appeal. The Supreme Court's decision in Condon will be important in terms of both legal principle and the day to day operations of the District Courts. But the decision either way will not be particularly relevant to the overall effectiveness of the Act in achieving its objectives.
  1. The primary issues which the Sentencing Act has thrown up for the courts have involved problems of implementation - not so much what the Act means but rather how it is to be applied. As well, in the present context, what is of primary importance is not what individual judges (or individual appellate courts) do in specific cases (eg as in Condon), but rather how the judiciary as a whole has responded to the Act.

The legislative policies underlying the Act

  1. The relevant policies of the Act include:

(a) Increasing sentences for the most serious offenders;

(b) Restating (although with some changes) the existing law and practice of sentencing.

  1. Given these two policies and the fiscal implications of any increase in overall rates of incarceration, there is a third relevant policy - to achieve the primary purposes of the Act with no more than a modest increase in the prison muster.

Increasing the sanctions imposed on the worst offenders

General

  1. The Act requires the courts to have more regard than previously to statutory maximum penalties (see s 8(c) and (d)) and provides in detail for the circumstances in which minimum periods of imprisonment are to be imposed in murder cases (see ss 102-104) and more generally for the imposition of minimum terms where determinate sentences have been imposed (see s 86). Section 86 was not in the Bill as originally introduced and was a response to those who were critical of the provisions of the companion parole legislation under which a prisoner would be eligible for (although not entitled) to parole at the expiry of one third of the nominal sentence.

The significance of maximum penalties

  1. I do not have any statistics which indicate, one way or the other, whether the enhanced relevance of statutory maxima has had significant impact on the lengths of prison sentences which are being imposed. What I can say, however, is that the new legislation has been recognised by the Court of Appeal in guideline sentencing judgments as a factor which must be recognised when tariffs are imposed.[4] This will have a filter down effect on sentencing decisions.

Sentencing for murder

  1. Under the new sentencing regime for murder:-

(a) There is no longer a mandatory life sentence. Instead there is a presumption in favour of life imprisonment, which is to be imposed unless it would be manifestly unjust given the circumstances of the offence and offender, see s 102(1).

(b) Where life imprisonment is imposed, there is a default non-parole period of 10 years.

(c) Extended non-parole periods may be imposed. Section 103 as originally enacted provided for this to occur if the court was satisfied that the circumstances of the case were "sufficiently serious to justify doing so", and the court might reach this conclusion if the circumstances took the offence "out of the ordinary range of offending of the particular kind". The minimum period imposed was required to be the minimum the court considers to be justified having regard to the circumstances of the case, including those of the offender. Section 103, as originally enacted was, in effect, a re-enactment of the corresponding provisions in s 80 of the Criminal Justice Act.

(d) Section 103 was simplified with effect from 1 July 2004. Under the recast s 103, the minimum sentence to be imposed is what the judge considers to be necessary to hold the offender accountable for the offence and for the purposes of denunciation and deterrence.

(e) There is a requirement to impose a 17 year minimum period of imprisonment (unless the court is "satisfied that it would be manifestly unjust to do so") in relation to the categories of murder specified in s 104.

  1. Although the last of the categories stipulated in s 104(i) ("any other exceptional circumstances") might suggest a legislative assumption that only a small percentage of murders would be caught by the section, a significant number (and perhaps most) involve (at least arguably) one or more of the aggravating features specified in s 104. For instance, virtually any premeditated and planned murder arguably involves "calculated ... planning" given that such planning need not necessarily be "lengthy" (see s 104(b)). Murder committed as a consequence of, or in association with, other offending will almost always be within s 104(a), (c) or (d). The murder of children in cases of domestic violence will almost necessarily fall within s 104(g).
  1. Here are the details of the non-parole periods imposed on offenders convicted of murder during the period 1995-2004:[5]
Non-parole period
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
10 years
23
26
31
24
20
21
14
16
10
11
>10 to 12 years
1
0
0
0
0
0
1
1
1
1
>12 to 14 years
2
3
3
0
2
5
4
4
5
2
>14 to 16 years
2
0
1
0
0
2
0
2
1
1
> 16 to 18 years
0
0
2
0
1
1
0
4
2
4
> 18 to 20 years
0
0
0
0
0
0
0
0
0
3
> 20 years
0
0
0
0
0
0
0
0
2
2
TOTAL
28
29
37
24
23
29
19
27
21
24
Overall average (years)
10.7
10.3
10.8
10.0
10.5
11.2
10.8
12.0
13.5
14.0
  1. In considering these figures it is important to recognise that many offenders sentenced in 2003 (and perhaps some who were sentenced in 2004) were not subject to the s 104 regime. My impression of practice is that s 104 is having the effect of increasing the number of cases in which minimum terms are imposed and increasing the lengths of such minimum terms. I suspect that rather more cases are being held to be within s 104 than was anticipated.
  1. The operation of the new regime has not been entirely smooth. The cases which have caused the greatest problems for the courts have involved young offenders[6] and those where the offender has pleaded guilty.[7] The ways in which the courts have addressed these (and indeed other) problems lie outside the scope of this paper. What the relevant cases have thrown up is the tightly prescribed nature of the discretions to deviate from the 17 year minimum in cases to which s 104 applies and to impose a sentence less than life (under s 102) in other cases. The courts may only do so if a manifest injustice test is satisfied. This necessarily produces a lumpy distribution of sentences and is thus not consistent with a seamless gradation of sentencing response to offender culpability. To put this in concrete terms, there are likely to be far more minimum terms of 17 years than of 16 years.

Minimum terms where determinate sentences are imposed

  1. Section 80(4)-(6) of the Criminal Justice Act provided a power to impose an extended non-parole period in respect of determinate sentences but only in cases of serious violence and to sentences of imprisonment over two years - that is, in cases where the offender was required under the then parole system to serve two thirds of the sentence. A characteristic of the cases decided under s 80(4)-(6) of the Criminal Justice Act was that the effect of the imposition of an extended non-parole period was usually limited and unlikely to add much more than 15% to the length of time the offender would actually have had to serve under the two-thirds rule. Given the rigidity of the two-thirds rule and the flexibility inherent in the sentencing process (and thus the range of acceptable outcomes), there was little point in invoking the section.
  1. Section 86 of the Sentencing Act permits a Judge who imposes a determinate sentence of imprisonment to fix a minimum non parole period not exceeding two thirds of the nominal sentence or 10 years, whichever is the lesser. Under the new parole regime, with parole eligibility arising at one third of the nominal sentence, s 86 operated in a context very different from that of the corresponding Criminal Justice Act provisions.
  1. Unsurprisingly, the courts ran into difficulties with this section almost immediately, see R v Brown.[8] It has been substantially amended. The amendments have undoubtedly improved the section but it remains problematical.
  1. The sentencing and parole legislation proceeds on the basis that in general an offender who is sentenced to more than two years imprisonment should be eligible for (although not entitled to) parole once one-third of the sentence has been served. In ordinary cases, the overall culpability of the offender is reflected in the nominal sentence imposed and thus in the minimum term which must be served (being one third of the nominal sentence).
  1. Where the nominal sentence to be imposed on the offender is at or near the maximum, the discretion to fix a longer than normal minimum term is unproblematic. In this context, the discretion provides an additional mechanism for distinguishing between the culpability of offenders. For instance, if two offenders each receive the maximum sentence of 14 years for attempted murder, the sentencing judge can make allowances for differences in individual culpability by fixing different minimum terms. Viewed in this way, the maximum sentence for attempted murder can be viewed as being 14 years imprisonment with a direction to serve two-thirds of that before parole is considered.
  1. Cases of the sort discussed in the preceding paragraph are comparatively rare and it is clear that the s 86 discretion was not intended to be reserved for such circumstances. But outside of those circumstances, the logic underpinning the section is elusive.
  1. A judge who intends to impose a particular nominal sentence (say six years) but who thinks a requirement to serve anything less than one half (say three years) is an insufficient response to the offending, can address that concern either by imposing a minimum term (in this instance three years) or by imposing a longer nominal sentence (on this arithmetic, up to nine years). A nine year sentence (carrying automatic parole eligibility after three years) is a more severe sentence than a six year sentence with a non-parole period fixed at three years. But a comparison between say a seven year sentence with no extended non-parole period fixed and a six year period with a three year non-parole period is not quite so easy. Depending on parole expectations, a six year sentence with a three year non-parole period may be a more severe sanction than a seven year sentence with no non-parole period fixed. Yet this seems incongruous because, in orthodox sentencing practice, the overall culpability of the offender is primarily reflected in the nominal sentence. In this hypothetical situation, the person whose overall culpability is lower (reflected in the six year sentence) is more harshly treated than the person whose culpability is higher (as reflected in the seven year sentence).
  1. Existing sentencing practice and tariff decisions mean that the fixing of nominal sentences is reasonably predictable, at least in the sort of cases which are likely to attract a minimum period of imprisonment. There is no corresponding set of principles for matching offender culpability and minimum terms.
  1. The courts have had difficulties identifying an appropriate basis for distinguishing between those cases in which a s 86 order is appropriate from those in which it is not. Some of these difficulties were discussed in R v Brown in which problems with s 86, as initially enacted, were pointed out and there have been other cases in which the Court of Appeal has sought to discern some underlying principle. But the extent to which those cases provide real guidance to sentencing judges is open to question.
  1. In a sense, the problems with s 86 are inherent in the controversial premise which underlies the Parole Act 2002, namely that the integrity of the sentence imposed by the sentencing judge is adequately maintained by a default requirement that offenders serve a minimum of a third of the sentence as imposed and that the balance, two-thirds, is in effect a discretionary component subject to the administration of the parole board.
  1. Minimum non-parole periods are imposed in comparatively few cases. Here is a breakdown of the sentences in which minimum periods of imprisonment were imposed between 1 July 2002 and 30 June 2003 in relation to determinate prison sentences, by type of offence:[9]
Offence Minimum term as a proportion of the total imposed sentence
 
40-49%
50%
51-59%
60-66%
67%
Total
Manslaughter
1
2
1
1
1
6
Attempted murder
3
0
2
0
1
6
Kidnap/abduct
0
0
2
0
2
4
Rape
0
10
9
7
7
33
Unlawful sexual connection
3
6
2
2
2
15
Aggravated robbery
0
5
1
4
6
16
Grievous assault
2
5
7
5
12
31
Other violence1
1
4
0
1
3
9
Property offence2
1
2
0
2
3
8
Deal in non-cannabis drugs
2
4
2
1
0
9
Other
0
1
1
0
0
2
Total
13
39
27
23
37
139
  1. A comparison between these figures and the statistics referred to later in this paper shows that minimum periods of imprisonment are imposed in a very small proportion of cases. It seems plausible to assume that the comparatively few who are subject to such orders will not usually be good candidates for release at or near parole eligibility date. Given this and what we know of the approach taken by the parole board to release decisions in respect of male offenders (also discussed later in the paper), most of the orders made under s 86 can have had little or no practical effect on the time the offenders concerned have spent (or will spend) in prison. On this basis, the intellectual effort devoted to, and the ink spilt on, the appropriate way to exercise the s 86 discretion has probably been largely wasted.

Congruence between the provisions of the Act and sentencing practice

  1. As noted, the Act - in large measure anyway - was intended to be a restatement of existing sentencing practice.[10] The view that the Act would have comparatively little impact on the prison muster must, in large measure, have been based on the assumption that this was so.
  1. In this context it is of significance that there are, however, at least one and possibly two respects in which the Act and current sentencing practice deviate.
  1. For any sentencing judge (and any appellate court reviewing a sentence), the concept of "starting point" now provides the primary method of measuring consistency. Ironically there was much inconsistency, at least up until recently, as to what the expression "starting point" refers to. Possibly for this reason and also perhaps because the concept is employed primarily where imprisonment is to be imposed, the expression simply does not feature in the Act.
  1. In the nearly four years which have elapsed since the Sentencing Act, there has been a growing consensus that the expression "starting point" should be used to refer to a notional sentence which would be appropriate for all those factors relevant to the culpability of the offence but excluding factors relevant to the offender and a plea of guilty, if any.[11]
  1. That the statute does clearly address the starting point approach to sentencing means that there is a major discrepancy between the sort of sentencing process which the Act appears to provide for and the daily practice of the Courts.
  1. Another fundamental element of current practice which is not addressed, at least explicitly, is tariff sentencing. The New Zealand origins of tariff judgments go back to the late 1970s. In R v Pawa[12] and R v Pui{13] sentences imposed for rape were reduced to bring them into line with the existing pattern of sentencing for sexual offences (a deviation from prior practice under which consistency was not a decisive consideration). Then in R v Smith{14] and R v Dutch[15] there were extensive reviews of existing appellate sentencing decisions. These are the first (or at least amongst the first) New Zealand tariff cases. In all four cases, the purpose of the exercise was to establish existing sentencing patterns. More recently, the Court of Appeal has adopted a more prescriptive "top-down" approach, as for instance in R v Mako.[16] By 2002, tariff sentencing was well developed and guideline judgments covered a significant proportion of the offences which were customarily dealt with in the High Court at least.
  1. The Sentencing Act refers to "consistency with appropriate sentencing levels"[17] but does not otherwise provide for tariff sentencing. The practice of appellate courts prescribing tariffs in the way in which the Court of Appeal has is not to everyone's taste,[18] and there was a feeling amongst a number of judges that the fact and offender specific nature of the Act was inconsistent with the continued authority of existing tariff cases. That in the end was not the approach which the courts took. Indeed the Court of Appeal, and perhaps more controversially the High Court, continue to provide tariff judgments.[19]
  1. A legal purist might be dismayed by the differences between actual sentencing practice and what appears to be contemplated by the Act. But these differences (and what they imply as to legislative understanding of the way sentencing is carried out) are of practical significance as well. This is because a legislative system which builds on existing judicial practice is unlikely to function as intended if there is any significant misunderstanding as to what that practice actually is. Such misunderstanding might account for some of the difficulties and perhaps anomalies which have arisen in respect of sentencing for murder and the fixing of minimum terms in relation to determinate sentences.

Implementation at a macro-level -; the impact of the Sentencing Act on current rates of incarceration

  1. One of the purposes of the Sentencing Act was to increase nominal sentence lengths (and sometimes minimum periods of imprisonment) for the most serious offending. But, that apart, the Sentencing Act largely built on and, indeed in a sense, partly codified existing mainstream sentencing practice. Given this, it was expected (or at least hoped) that the impact of the Act on the prison muster would be limited.
  1. What effect has the Act had on prison musters?
  1. Here are the inmate and home detention figures in biennial blocks between 1991-2003 :[20]
Year
Sentenced females
Sentenced males
Remand females
Remand males
Home detention
Total
1991
139
3682
9
402
0
4232
1993
118
3645
14
486
0
4263
1995
151
3981
14
343
0
4489
1997
207
4728
13
516
0
5464
1999
206
4759
24
633
25
5647
2001
202
4716
32
830
174
5954
2003
262
4833
73
1072
595
6835

It should be noted that only the last 18 months of these figures can have been affected by the Act. Over the period covered, inmate numbers increased as follows:

Year
1991-1993
1993-1995
1995-1997
1997-1999
1999-2001
2001-2003
Increase
0.7%
5.3%
21.7%
3.3%
5.4%
14.7%
  1. The Corrections website records the number of inmates (presumably including those on home detention) as 7,565 as at 7 February 2006 of whom 5,940 were sentenced prisoners and 1,552 were on remand (figures which, incidentally do not tally).
  1. Current forecasts predict significant increase in our prison musters up to 2010.[21]
  1. All of these figures might suggest that the Act has tended to drive up the prison muster. But the reality seems to me to be more complex.
  1. A number of the drivers of current and forecast incarceration rates are not related to the introduction of the Sentencing Act. In this category are prosecution and conviction rates (which for key offences are increasing), the extent to which defendants awaiting trial are remanded in custody and parole board decisions as to release on home detention and parole.[22]
  1. At this point, a brief mention of the parole system is appropriate.
  1. Prisoners serving sentences of two years or less serve half their sentences. Those serving more than two years are eligible for parole after serving one third of the sentenced but may be detained in prison for the entire duration of the sentence.
  1. My impression from the material I have seen is that male prisoners serving sentences in excess of two years can expect to serve 50% or more of the nominal sentence and that a significant proportion (perhaps 12% or more) will serve the entire sentence. In the case of those sentenced in 2003 the pattern for male prisoners has been that approximately one quarter serve 33%, five eighths serve 50% and one eighth serve 100% of the sentences imposed. The average proportion served was 52%. On the other hand, five eighths of female prisoners serve 33%, one quarter serve 50% and one eighth serve 100% of the nominal sentence. The average length of sentence served by women was 42%.[23]
  1. If the Sentencing Act has had the effect of increasing prison musters, this is presumably because it has caused or contributed to increases in either the rate at which imprisonment is imposed on offenders or the average length of sentences imposed.
  1. The percentage of defendants who are sentenced to imprisonment is greater than was the case prior to the Sentencing Act, in the case of males by about 5.5% over the last three years and, in the case of females by 53%.[24] This latter, and apparently dramatic development (which must raise many questions) is not particularly significant in terms of the overall prison muster as comparatively small numbers of women are imprisoned.
  1. The length of prison sentences increased in 2002 and 2003 but reduced in 2004. In 2005, sentences were in line with averages for the period 1999-2004.[25] So the Sentencing Act does not appear to have had an effect (at least one which is currently able to be discerned) of increasing average sentence length.
  1. Overall, my impression is that the introduction of the Sentencing Act in 2002 has not been the primary driver of increases in the prison muster. To put it another way, if the Sentencing Act had not been passed, the prison muster would not be much different from what it is currently.

Conclusions - or drawing the threads together

  1. Given the limitations of the statistical material available and the comparatively short time which has elapsed since the Act came into effect, any conclusions as to its implementation must be tentative. Nonetheless, the emerging material is at least of interest. And there are three comments which can be made with reasonable confidence:

(a) There has been (to date anyway) reasonable concordance between the intended impact of the Act and its actual consequences; in particular sentencing has become more severe for the most serious offenders but the courts have resorted to prison as a sanction only slightly more than previously with average lengths of sentence not much affected.

(b) Many of the difficulties which the courts (and particularly appellate courts) have encountered are associated with the default entitlement to parole at the expiry of one third of sentence (in the case of sentences over two years). The controversy as to this automatic eligibility resulted in the enactment of the problematical s 86 which, despite subsequent improvement, continues to prove difficult to apply. Allowing for parole board practice, the debate about parole eligibility dates for male prisoners and the application of s 86 now seem to be largely beside the point.

(c) Although the Act set out to restate sentencing practice, it does not engage (at least substantially) with two significant features of that practice, the concept of starting point and guideline or tariff judgments. This implies an incomplete legislative understanding of the pre-existing judicial practice. Interestingly, the key areas of difficulty for the courts in their implementation of the Act have been associated with a failure (at least in the case of s 86 and possibly in respect of sentencing for murder) to think through the practical implications of legislative changes on the way that the courts would carry out sentencing exercises.

Footnotes

1 President of the Court of Appeal.
2 [1992] 1 NZLR 555.
3 [2005] 1 NZLR 446.
4 See, for instance, R v Taueki [2005] 3 NZLR 372.
5 Figures supplied by the Ministry of Justice to the Court of Appeal.
6 See for instances R v Slade [2005] 2 NZLR 562 and R v Green CA461/04 2 June 2005.
7 See R v Williams [2005] 2 NZLR 506 and R v Walsh (2005) 21 CRNZ 946.
8 [2002] 3 NZLR 670.
9 Rajesh Chhana, Philip Spier, Susan Roberts, Chris Hurd “The Sentencing Act 2002 : Monitoring the First Year” (A paper prepared by the Ministry of Justice, 2004).
10 See also Warren Young and Neville Trendle, “Developing the Sentencing Framework: The Sentencing Act and Beyond” in Essays on Criminal Law: A Tribute to Professor Gerald Orchard. 
11 See for instance R v Taueki [2005] 3 NZLR 372 at [42] – [44].
12 [1978] 2 NZLR 190.
13 [1978] 2 NZLR 193.
14 [1980] 1 NZLR 412.
15 [1981] 1 NZLR 304.
16 [2000] 2 NZLR 170.
17 In s 8(e).
18 See for instance R v Makarian (2005) 79 ALJR 1048.
19 Taueki is an example.
20 David Harpham Census of Prison Inmates and Home Detainees 2003 (A paper prepared by the Department of Corrections, 2004).
21 See Gray and Corbu, “Prison Population Forecast 2005” (A paper prepared by the Ministry of Justice). 
22 See Gray and Corbu.
23 See Gray and Corbu.
24 See Gray and Corbu.
25 See Gray and Corbu.

 

 

This page was last modified on