VALENTINE ATIRIANA MOSES v R [2020] NZCA 296

VALENTINE ATIRIANA MOSES v R [2020] NZCA 296 [15 July 2020]IN THE COURT OF APPEAL OF NEW ZEALANDI TE KŌTI PĪRA O AOTEAROACA167/2019[2020] NZCA 296BETWEEN VALENTINE ATIRIANA MOSESAppellantAND THE QUEENRespondentHearing: 30 April 2020Court: Kós P, Miller and Collins JJCounsel: V L Thorpe and A W Clarke for AppellantM J Lillico and ZWQ Andrew for RespondentE A Hall and L Scott for Criminal Bar Association of NewZealand as IntervenerJudgment: 15 July 2020 at 11.00 amJUDGMENT OF THE COURTA The application to adduce a s 27 report on appeal is granted.B The appeal is allowed.C We quash the sentence passed below and substitute a sentence of sevenyears, five months imprisonment.____________________________________________________________________REASONS OF THE COURT(Given by Miller J)TABLE OF CONTENTSThe three-step methodology [4]The Sentencing Act framework for sentence calculation [4]The starting point: R v Taueki [5]Uplifts and discounts must maintain proportionality [8]What the Act says about a guilty plea discount [11]The rationale and three-step methodology: Hessell v R [12]Calculating the guilty plea discount [27]Other mitigating factors reduce the guilty plea deduction [30]The three-step methodology in practice [39]Should the Court depart from Hessell? [44]A modified methodology for calculating guilty plea discounts [45]Ms Moses's sentence appeal [50]The facts [50]The sentencing [54]The appeal [59]The starting point [60]Personal mitigating factors: the s 27 report [63]Sentence calculation [71]Result [72][1] In this judgment we revisit the Court's 2009 judgment in Hessell v R, whichadopted a three-step methodology for sentencings affected by a guilty plea.1 We alterthat methodology to require a two-step approach in which any discount for a guiltyplea is fixed at the second step.[2] Ms Moses is one of several dealers in methamphetamine who were sentencedunder the former guideline judgment, R v Fatu,2 and whose appeals were adjournedpending delivery of the Court's current guideline judgment in Zhang v R.3 We alsoconsider, by reference to the Zhang guidelines, whether her sentence was manifestlyexcessive.41 Hessell v R [2009] NZCA 450, [2010] 2 NZLR 298 [Hessell (CA)]. The three-step methodologysurvived the Supreme Court judgment: Hessell v R [2010] NZSC 135, [2011] 1 NZLR 607[Hessell (SC)] at [73].2 R v Fatu [2006] 2 NZLR 72 (CA).3 Zhang v R [2019] NZCA 507, [2019] 3 NZLR 648.4 As with other appeals brought before Zhang v R was delivered, we approach the appeal not as asearch for error in the court below but by assessing the sentence against the Zhang guidelines.[3] We address the methodological issue before turning to the circumstances ofMs Moses's appeal.The three-step methodologyThe Sentencing Act framework for sentence calculation[4] The Sentencing Act 2002 establishes the framework within which this Courthas developed sentencing methodology. The Act records purposes for which offendersmay be sentenced and lists principles and aggravating and mitigating factors thatcourts must consider in sentencing.5 It does not rank sentencing purposes or principlesor prescribe that any given aggravating or mitigating factor must be given more or lessweight than another.6 Its list of aggravating and mitigating factors is not exhaustive.7Nor does it adopt any particular methodology. But it does invite a structured approachin which the sentencer evaluates the seriousness of the harm done, the culpability ofthe offender's conduct, the interests of the victim, and the personal circumstances ofthe offender. Guideline judgments build on the Act, seeking to further its purposes bypromoting transparency of analysis, which facilitates comparison and appellatereview, and principled consistency of outcome.The starting point: R v Taueki[5] An account of sentencing methodology under the Act begins with R v Taueki,a 2005 decision in which this Court reviewed longstanding guidelines for grievousbodily harm offences and set increased starting points for serious offences.8The former guidelines had established an array of sentences without making it clearwhether they incorporated aggravating and mitigating circumstances referable to theoffender.9[6] In Taueki the Court accordingly established a two-step methodology foundedon the idea of a starting point. At the first step the sentencing judge establishes a5 Sentencing Act 2002, ss 7, 8 and 9.6 Hessell (SC), above n 1, at [37].7 Section 9(4).8 R v Taueki [2005] 3 NZLR 372 (CA).9 At [8] citing R v Hereora [1986] 2 NZLR 164 (CA).provisional sentence or starting point based on the circumstances of the offending.10The Court defined the starting point as the sentence considered appropriate for theparticular offending by an adult offender after a defended trial.11 We use the term"adjusted starting point" to signify that it incorporates all aggravating and mitigatingfeatures of the offending.12 At the second step the judge tailors the adjusted startingpoint to the offender, incorporating their personal aggravating and mitigatingcircumstances to reach the appropriate end sentence.13 By separating thecircumstances of the offence from those of the offender the methodology permitscomparison among starting points for similar offending.14[7] The Court has since used adjusted starting points to demarcate sentencingbands in other guideline judgments. Apart from Taueki, it has done so in R v AM(sexual violation) and Zhang v R (methamphetamine).15 It has also deliveredjudgments providing guidance about some, but by no means all, of the mitigatingfactors recognised in the Act.16Uplifts and discounts must maintain proportionality[8] The Act does not treat proportionality as a dominant sentencing principle, butits purposes, principles, and aggravating and mitigating factors emphasise harm andculpability and by requiring that the sentencer consider these matters the Act doesanticipate that the sentence should be commensurate.17 By way of illustration, s 8(a)requires that the sentencer take into account gravity of the offending, while ss 8(b)–10 R v Taueki, above n 8, at [28]–[30].11 At [8] following R v Mako [2000] 2 NZLR 170 (CA) at [34].12 By way of illustration, for the grievous bodily harm offending addressed in R v Taueki mitigatingfeatures of the offence included provocation and excessive self-defence: at [32].13 At [44].14 At [43]; and Zhang v R, above n 3, at [134].15 R v AM (CA27/2009) [2010] NZCA 114, [2010] 2 NZLR 750; and Zhang v R, above n 3.16 Apart from Hessell, the Court has addressed youth (Churchward v R [2011] NZCA 531, (2011) 25CRNZ 446), mental capacity (E (CA689/10) v R [2011] NZCA 13, (2011) 25 CRNZ 411; andShailer v R [2017] NZCA 38, [2017] 2 NZLR 629), addiction and mental health (Zhang v R, aboven 3), and ill-health (R v Verschaffelt [2002] 3 NZLR 772 (CA)).17 See Mirko Bagaric "Injecting Content into the Mirage that is Proportionality in Sentencing"(2013) 25 NZULR 411 at 416–417; Richard G Fox "The Meaning of Proportionality inSentencing" (1994) 19 Melb UL Rev 489 at 492–493; and Andrew Ashworth "Re-evaluating theJustifications for Aggravation and Mitigation at Sentencing" in Julian V Roberts (ed) Mitigationand Aggravation at Sentencing (Cambridge University Press, Cambridge, 2011) 21 at 25.(d) require that the offending be gauged against maximum penalties,18 and s 8(g)requires that the sentence be the least restrictive that is appropriate in thecircumstances. Some sentencing purposes, principles and factors do not rest onproportionality — the leading examples are public protection and rehabilitation — butit is always an important sentencing principle. It is underscored by s 9 of theNew Zealand Bill of Rights Act 1990, which protects the right not to be subjected todisproportionately severe punishment.[9] The authorities recognise that uplifts and discounts for personal circumstancesshould also be proportional. This principle is seen most clearly in cases about upliftsfor previous convictions, but it also applies to discounts.19 So, for example, inTaylor v R this Court held that an uplift "must bear some reasonable relationship orproportionality to the starting point".20 This ensures that end sentences retain anappropriate degree of proportionality, in the offender's circumstances, to the harmdone and the culpability of the offending.[10] To say that uplifts or discounts at the second step must bear an appropriateproportion to the adjusted starting point is not to insist that they be calculated as apercentage.21 Judges commonly fix some discounts as a number of months or years.But it is usual to fix larger discounts, including those for youth, mental illness andguilty pleas, as a percentage, which aids calculation and facilitates comparison amongcases.22What the Act says about a guilty plea discount[11] Section 9(2) of the Act provides that a sentencing court must considermitigating factors listed there, to the extent they apply to the case. One of these is18 See Andrew von Hirsch "Proportionate Sentences: A Desert Perspective" in Andrew von Hirsch,Andrew Ashworth and Julian Roberts (eds) Principled Sentencing: Readings on Theory andPolicy (3rd ed, Hart Publishing, Oxford, 2009) 115 at 118.19 R v Fahey CA184/00, 2 November 2000 at [33]; and Orchard v R [2019] NZCA 529 at [41].20 Taylor v R [2012] NZCA 332 at [46]. See also Tiplady-Koroheke v R [2012] NZCA 477 at [24];Blackmore v R [2014] NZCA 109 at [13]; and Hodgkinson v R [2012] NZCA 478 at [21].21 Blackmore v R, above n 20, at [13] n 18.22 See for example Whittaker v R [2020] NZCA 241 at [49]."whether and when the offender pleaded guilty".23 The statutory language saysnothing about the rationale for a guilty plea discount, or its amount.The rationale and three-step methodology: Hessell v R[12] Until 2009, when this Court's judgment in Hessell was delivered, the leadingauthority on guilty plea discounts was R v Mako, a 2000 decision in which the Courtresisted laying down any specific amount or proportion because of the widely varyingcircumstances in which a plea might be entered.24 Following the Sentencing Act 2002a practice had nonetheless developed of applying a sliding scale based on the plea'stimeliness. In Hessell the Court abandoned the Mako approach and established a setof guidelines designed to deliver a consistent approach to guilty plea discounts.25[13] The decision resulted in a three-step methodology for sentences affected by aguilty plea. As explained above, at the first step the court fixes an adjusted startingpoint based on the circumstances of the offence, and at the second step it incorporatesany aggravating and mitigating circumstances of the offender. Hessell added the thirdstep, which provides for the guilty plea.26 The Court established a sliding scale whichpermitted a discount of 33 per cent for a plea entered at first reasonable opportunity,reducing to 10 per cent for a plea entered three weeks before trial.27 It recorded thatits approach was based on its then-current practice.28[14] The Court adopted the three-step methodology because it identified a strongpublic interest in guilty plea discounts being predictable. It cited guidelines preparedby the English Sentencing Guidelines Council and a draft guideline prepared by theNew Zealand Law Commission in the expectation that a sentencing council would beestablished in New Zealand. These guidelines justified a guilty plea discount becauseit avoids a trial, with associated benefits for victims and witnesses and efficiency gains23 Section 9(2)(b).24 R v Mako, above n 11, at [14].25 Hessell (CA), above n 1, at [6].26 At [14].27 At [15].28 At [14]. See R v Hannaghan CA396/04, 9 June 2005 at [25]–[26]; R v Growden CA67/05,25 October 2005 at [50]; R v Fonotia [2007] NZCA 188, [2007] 3 NZLR 338 at [50]; R v Andersen[2007] NZCA 288 at [35]; and R v Proctor [2007] NZCA 289 at [27].for the State; for that reason, it was distinguished from personal mitigation.29The English guideline accordingly recommended that "the sentencer should addressseparately the issue of remorse, together with any other mitigating features, beforecalculating the reduction for the guilty plea".30 The discount should be a proportionof the total sentence imposed, calculated according to "the circumstances in which theguilty plea was indicated".31 Under the Court of Appeal's approach the amount of thediscount did not depend on the strength of the prosecution case. The underlyingjustification is that a guilty plea is an autonomous decision to surrender the right toproof.32 Under this approach the discount should not be reduced because the offenderis thought to have had little chance of acquittal on the facts that he or she has admittedwith the plea.[15] Although the Court followed the English methodology by separating the guiltyplea from other mitigating factors, it departed when setting the maximum amount ofthe discount and in its treatment of remorse. Under the English approach a guilty pleaearns a maximum discount of one-third or 33 per cent and remorse is assessedseparately, at the second step. There is no guideline, but a genuinely remorsefuldefendant who pleads guilty at the earliest opportunity may receive a material discountin addition to the one-third available for the guilty plea.33 In Hessell this Court heldthat the maximum discount of 33 per cent included remorse, for which an additionalallowance might be made only in exceptional cases where it had been demonstrated ina practical and material way.34 The Court justified bundling the guilty plea withnon-exceptional remorse on four grounds: a guilty plea is the best evidence of remorse;an allowance for remorse is "automatically built in" to the guilty plea discount;29 United Kingdom Sentencing Guidelines Council Reduction in Sentence for a Guilty Plea:Definitive Guideline, Revised 2007 (23 July 2007); and New Zealand Law CommissionSentencing Establishment Unit Draft Sentencing Guidelines on Discount for Guilty Plea(July 2008) at [4].30 Definitive Guideline Revised 2007, above n 29, at [2.4]. Any allowance for assistance to theauthorities was to be calculated separately in the same way.31 [At 4.1]. The current (2017) English guidelines continue to adopt this approach: United KingdomSentencing Council Reduction in Sentence for a Guilty plea: Definitive Guideline (1 June 2017)at 5.32 The literature is discussed by Richard Nobles and David Schiff in "Criminal Justice Unhinged:the Challenge of Guilty Pleas" (2019) 39 Oxford J Legal Stud 100. The 2007 United KingdomGuidelines, above n 29, allowed the strength of the prosecution case to be taken into account inexceptional circumstances. The 2017 version, above n 31, specifies that the benefits of a guiltyplea apply regardless of the strength of the evidence against the offender.33 See by way of example R v Pitts [2014] EWCA Crim 1615 at [63]–[65].34 Hessell (CA), above n 1, at [28].remorse is easily claimed but not easily gainsaid; and the guilty plea discount wouldbe more predictable if it incorporated remorse.35[16] Hessell went on appeal to the Supreme Court, which substantially modifiedthis Court's approach, beginning with the rationales for a guilty plea discount.The Supreme Court observed that the Act treats a guilty plea as a mitigating factor andindicates that an earlier plea should generally carry more weight than a later one, butotherwise indicates no policy governing the approach sentencing judges should take.36The Court accordingly turned to sentencing law developed by the courts, noting thatmuch of it had found its way into the Act's purposes, principles and factors.37The Court cited R v Strickland, a 1989 judgment in which this Court identified threepolicy rationales for a guilty plea discount:38 it spares the victim the ordeal of giving evidence; it saves the State the timeand expense of a defended hearing; and it may be evidence of the offender'sacceptance of responsibility for wrongdoing and contrition.[17] The Supreme Court confirmed that benefits to the criminal justice system, andto participants in it, supply the "core justification" for guilty plea discounts.39Apart from cost savings for the State and reduced trial backlogs, there are socialbenefits; witnesses and especially victims are spared the burden of giving evidence,and a guilty plea "often also assists victims and their families through itsacknowledgement of responsibility for the offending".40 We observe that this lastrationale values contrition, manifested through the guilty plea, not for itself butbecause, and to the extent that, atonement benefits the victim.[18] The Supreme Court held that this Court had erred by bundling the guilty pleadiscount with remorse, which must be treated as a discrete mitigating factor,41 andplacing too much importance on consistency of outcomes. Addressing this Court'sconcern about proof of remorse, the Supreme Court observed that the Act envisages itmust be "shown", meaning that sentencing judges need not take unsubstantiated35 At [24]–[28].36 Hessell (SC), above n 1, at [50].37 At [39]–[42].38 At [28] citing R v Strickland [1989] 3 NZLR 47 (CA) at 51.39 At [46].40 At [45].41 At [64].claims at face value.42 A guilty plea may show only that the defendant knows theprosecution case is strong.43[19] For these reasons the Supreme Court rejected this Court's scaled discountapproach, holding rather that a guilty plea discount requires an evaluative assessmentreflecting all the circumstances of the case, including the strength of the prosecutioncase and the point at which the defendant had the opportunity to be informed of allimplications of the plea.44 It follows that an early plea need not earn a full discount.[20] The Supreme Court capped the guilty plea discount at 25 per cent.45 Althoughit expressed concern that this Court's prescriptive approach, narrowly focused ontiming of the plea, might induce the innocent to plead guilty, the Court also noted therewas no evidence that New Zealand courts' traditionally flexible approach, under whichdiscounts of up to one-third were given for an early plea, had led to perverseoutcomes.46 This is not the occasion to revisit policies underpinning the upper limitto a guilty plea discount. That exercise would call for information about the systemicbenefits of a guilty plea, which may vary with circumstances affecting the courtsystem, analysis of the risk that too large a discount may create too powerful anincentive to plead guilty, and review of processes that sentencing courts might employto manage that risk.[21] However, the cap does make it necessary to decide which circumstances maycontribute to the guilty plea discount and which may justify an additional discount forremorse. The Supreme Court did not focus on that question. Its judgment is open tothe interpretation that remorse may contribute substantially to the guilty pleadiscount,47 and if that is so, there may be little room for an additional remorse discount.This Court subsequently interpreted Hessell in that way, holding in R v Clifford thatonly remorse that can be characterised as "extraordinary" warrants an additional42 At [64].43 At [62].44 At [57], [65] and [74]–[77].45 At [75].46 See [71]–[72]; R v Walker [2009] NZCA 56 at [19]; and cases cited at n 28 above. We observethat the Law Commission had proposed a 33 per cent cap.47 When rejecting this Court's view that strength of the prosecution case is irrelevant, the SupremeCourt stated (at [60]) that acceptance of responsibility is an important factor.discount.48 We prefer the view that, when its judgment is read as a whole, the SupremeCourt did not mean to go so far. We make four points.[22] First, benefits to the judicial system and participants in it supply the principaljustification for a guilty plea discount. To this extent New Zealand practice isconsistent with the rationale for guilty plea discounts adopted in sentencing guidelinesproposed by the Law Commission and adopted in English law. In addition, atonementexperienced by the victim following the offender's remorse may now justify adiscount, though we observe that such cases may be uncommon in practice.The sentencing judge must decide which of these rationales applies and what weightwill be given to them.[23] Second, fixing the amount of the discount requires an evaluative judgment, butthe relevant circumstances of the case must be those that engage any applicablerationales for the discount. The rationales established by the Supreme Court suggestthat, among other things, the scale and complexity of the trial, the proximity of theplea to first appearance or to trial,49 the justification for any delay, the inevitability orotherwise of conviction, the benefits of not giving evidence for victims and witnesses,and the victim's experience of atonement following the offender's acceptance ofresponsibility may affect the amount of the discount, which may range from25 per cent to nothing.[24] Third, remorse is a personal mitigating factor that may justify a discountseparately from any guilty plea discount. Remorse is a question of fact and judgement.The defendant bears the onus of showing that it is genuine, meaning that it qualifiesas remorse and he or she actually experiences it.50 Remorse need not be extraordinaryto earn a discount, but it does require something more than the bare acceptance of48 R v Clifford [2011] NZCA 360, [2012] 1 NZLR 23 at [60].49 The earliest reasonable opportunity is a function of procedures for charging, appearances anddisclosure. In Hessell (CA), above n 1, which was decided before the Criminal Procedure Act2011 was enacted, this Court considered that a plea at second appearance would generally beconsidered early: at [29]. The Supreme Court referred more generally to the point at which thedefendant had the opportunity to be informed of all implications of the plea: Hessell (SC), aboven 1, at [75].50 Sentencing Act, s 24(2)(d).responsibility inherent in the plea. Courts look for tangible evidence, such asengagement in restorative justice processes.[25] Fourth, a guilty plea is not synonymous with remorse but may evidence it.The plea is an act of confession to a wrong done, and it is commonly associated withcontrition and a desire for expiation.51 It follows that guilty plea and remorse discountsmay be paired, and very often are: a defendant who pleads guilty at the earliestreasonable opportunity may also earn a remorse discount, while another who delaysthe plea until arraignment at trial may be denied both discounts.[26] We observe that, following Hessell, remorse may both justify its own discountand underpin a guilty plea discount where the latter is based on victims' experience ofatonement. In addition, s 10 of the Sentencing Act requires that a sentencing courttake into account an offer of amends, considering whether the offer is genuine andwhether the victim has accepted it as expiating or mitigating the wrong. It followsthat in those cases where victim atonement is a factor, and especially where a guiltyplea is accompanied by an offer of amends, it may sometimes be necessary to ensurethe sentencing analysis does not result in remorse receiving too much or too little creditin the end sentence.Calculating the guilty plea discount[27] That brings us to how the discount works. In Hessell this Court held that itshould be calculated as a percentage of the sentence that would otherwise have beenpassed; that is, after making allowances for aggravating and mitigating factors relatingto the offender:[14] we recommend to trial judges, as a general practice, that theyrecognise the guilty plea by giving a discrete reduction to what the sentencewould otherwise have been, such reduction to be calculated as a proportion ofthe total sentence that would otherwise be imposed. The extent of thereduction should depend on the stage in the proceedings at which a guilty pleais entered or at which the offender expresses a willingness to enter a guiltyplea to the offence for which he or she is later convicted. The reduction shouldbe made as the final step in the sentencing process; that is, after theappropriate sentence has been determined with reference to aggravating and51 Brendon Murphy "The Technology of Guilt" (2019) 44 Australasian Journal of Legal Philosophy64 at 88–89.mitigating factors relating to the offence and aggravating and all othermitigating factors relating to the offender. (Emphasis added)[28] The Supreme Court was at pains to emphasise that sentencing is an evaluativeexercise, but it chose not to adopt the Australian "instinctive synthesis" approach.Instead it adhered to what it described as the flexible New Zealand practice ofaddressing "gravity and culpability of offending" as "separate matters".52 The Courtcautiously went on to accept the three-step methodology:[73] There is no objection in principle to the application of a reduction ina sentence for a guilty plea once all other relevant matters have been evaluatedand a provisional sentence reflecting them has been decided on. Indeed, thereare advantages in addressing the guilty plea at this stage of the process (alongwith any special assistance given by the defendant to the authorities). It willbe clear that the defendant is getting credit for the plea and what that credit is.This transparency validates the honesty of the system and provides a degreeof predictability which will assist counsel in advising persons charged whohave in mind pleading guilty.[29] Mr Lillico argued that in this passage the Supreme Court endorsed the practiceof fixing the guilty plea discount as a percentage of the notional sentence calculated atstep 2. We do not agree. The Court held only that there is no objection in principle toa three-step approach in which the guilty plea discount is deducted as the last step.It approved the use of a percentage discount but it did not say that the three-stepapproach is mandatory. It did not consider whether the multiplicand (the figure towhich the percentage is applied) ought to be the product of step 1 or step 2.As Mr Lillico accepted, there is no reason to think the issue now before us was drawnto the Supreme Court's attention. On Hessell's facts it did not matter whether thediscount was taken as a percentage of the sentence provisionally assessed at step 1 orstep 2.53Other mitigating factors reduce the guilty plea deduction[30] Under the three-step methodology uplifts or discounts for personalcircumstances at step 2 affect the amount of the guilty plea deduction. The greater the52 Hessell (SC), above n 1, at [55]. The Court had earlier spoken approvingly, at [26] and [31], ofthe practice of issuing guideline judgments. See also R v Clifford, above n 48, at [51].53 The sentencing Judge had adopted an orthodox starting point based on the offence and added onemonth for personal aggravating factors before allowing a guilty plea deduction "in the region of10 per cent": Hessell (CA), above n 1, at [89]. The guilty plea was the only mitigating factor, andthe amount of the discount evidently was not affected by the small uplift at the second step.step 2 discounts for personal mitigating factors the lower the resulting multiplicandand the fewer units of time deducted for the same percentage guilty plea discount, asthe table below illustrates. We have assumed a 10-year adjusted starting point, a15 per cent discount for personal mitigating factors and a full guilty plea discount inthis hypothetical case:Three-step methodology Two-step methodologyStep 1 Begin with adjusted starting point120 months (10 years)Begin with adjusted starting point120 months (10 years)Step 2 Discount for personal mitigatingfactors (15%)120 – 120 x 15%= 120 – 18= 102Discount for SUM of personalmitigating factors (15%) andguilty plea discount (25%)120 – 120 x (15% + 25%)= 120 – (18 + 30*)= 120 – 48= 72Step 3 Discount for guilty plea (25%)102 – 102 x 25%= 102 – 25.5*= 76.5Endsentence6 years, 4 ½ months 6 years* The bolded figures show the number of months being deducted for the 25 per centguilty plea discount.[31] The Crown has not suggested any justification for reducing a guilty pleadeduction — the units of time credited — when the plea is accompanied by personalmitigating factors. A guilty plea discount in New Zealand now rests principally on thesystemic and social rationales we have discussed above — savings to the State, relieffor victims and witnesses from the burden of giving evidence, and victims' experienceof atonement following the offender's contrition. These rationales are likely to justifyaltering the amount of the discount according to the timing of the plea. But none ofthem justifies systematically reducing the deduction merely because an offender'ssentence is also to be discounted for other mitigating factors.[32] Mr Lillico argued rather that the three-step methodology makes no differencein the end. To begin with, uplifts for previous convictions and the like are usuallymodest, and in many cases the guilty plea supplies the only significant mitigatingfactor. In such cases it is usually immaterial whether a guilty plea discount is appliedbefore or after allowing for personal circumstances. We accept that submission holdstrue in many cases.[33] Mr Lillico next argued under either approach the effective sentence will stillfall comfortably within a range appropriate to the offence and the offender.He submitted that in such cases an appellate court is unlikely to find the sentencemanifestly excessive. We accept that submission as a general proposition, but ofcourse we are concerned here with a question of methodology which is of generalapplication to sentences affected by a guilty plea.[34] Finally, Mr Lillico argued that in practice judges use their discretion to ensurethe end sentence is just. We accept that submission for several reasons. First, we haveexplained that the Sentencing Act catalogues purposes, principles and factors withoutranking or weighting them and, as the Supreme Court emphasised in Hessell,sentencing is an evaluative exercise. Guideline judgments structure the exercise ofdiscretion but do not eliminate it; every guideline judgment recognises that judgesmust apply the Act and may depart from the guidelines where appropriate.54This Court's judgment in Hessell similarly recorded that the guidelines establishedthere did not override the discretion of sentencing judges.55 The only concrete limit isthat established by the cap of 25 per cent on the guilty plea discount.[35] Second, those guideline judgments which establish sentencing bands usuallydo so by reference to multiple considerations and leave it to the sentencing judge toestablish their respective relevance and weight in any given case. In R v AM, forexample, the Court listed 13 culpability factors for sexual violation offences withoutprioritising among them.56 Further, offenders are frequently sentenced for more than54 See R v Fatu, above n 2, at [32]–[36] in which this Court recognised that offending could falloutside of the prescribed bands. See also R v Taueki, above n 8, at [10]; R v AM, above n 15, at[36], [79] and [83]–[84]; and Zhang v R, above n 3, at [48] and [120].55 Hessell (CA), above n 1, at [6].56 R v AM, above n 15, at [37]–[64].one offence and that introduces the totality principle, for which there is no guidelinejudgment.[36] Third, there is no guideline for many aggravating or mitigating factors, andsentencing judges are left to gauge their relevance and weight in any given case.57[37] Finally, guideline judgments emphasise that the sentencing judge should standback and inquire whether the final sentence is correct in all the circumstances.In Hessell the Supreme Court affirmed this point:[77] All these considerations call for evaluation by the sentencing judgewho, in the end, must stand back and decide whether the outcome of theprocess followed is the right sentence.We observe that this Court, and the High Court, have often emphasised this approachto sentencing after considering the effect of the methodology on the guilty pleadiscount.58[38] All of that said, the three-step methodology is logically capable of affectingoutcomes in practice, as we have recognised in the table above. The examples towhich we next turn suggest that it has had that effect in some cases where the startingpoint was set under a guideline judgment.The three-step methodology in practice[39] We begin with Reweti v R,59 an appeal from a sentence of two years andnine months imprisonment for aggravated robbery, for which the guideline judgmentis R v Mako.60 Simon France J remarked that the starting point of four years had beenuncontroversial at first instance because the circumstances largely reflected one of theexamples given in Mako. The appellant was 18, and although he had some criminal57 But see cases listed above at n 16.58 The two-step approach was adopted in Reweti v R [2018] NZHC 809; R v Kokiri [2019] NZHC501; Barlow v R [2019] NZHC 725; and Royal v R [2020] NZCA 129. Its availability was alsocommented on in R (CA217/2018) v R [2018] NZCA 582; Peke-Meihana v R [2019] NZHC 642;Houkamau v Police [2019] NZHC 2743; Jones v R [2019] NZHC 1816; Solicitor-General vKaokao [2019] NZHC 2352; Harris v Police [2019] NZHC 2846; and Edmonds v Police [2019]NZHC 3038.59 Reweti v R, above n 58.60 R v Mako, above n 11.history he was an excellent candidate for rehabilitation. He also experienced ill-health.The Judge decided on a discount of 30 per cent for these mitigating factors and25 per cent for his early plea. Under the three-step methodology the end sentence was25 months, three months higher than it would have been if the guilty plea discount wasapplied to the adjusted starting point. To achieve what he found to be the right result,Simon France J took the latter approach, resulting in a sentence that could be, and was,converted to home detention.61[40] In Peke-Meihana v R, also an aggravated robbery appeal, Mallon J similarlyreached an end sentence of home detention although she did so by adjusting thestarting point rather than altering the guilty plea discount.62 Had she retained theoriginal starting point the alternative two-step methodology would have resulted in adifference of 10 months in the end sentence.63 The case demonstrates that themethodology can make a significant difference to the end sentence when othermitigating factors are very substantial.[41] We observe that the outcomes in these cases were available under Hessell.The Court of Appeal recommended its methodology to trial judges as a generalpractice and observed that the methodology did not readily accommodate alternativesentences:[52] Sometimes it may be appropriate to recognise a guilty plea byimposing one type of sentence rather than another. For example, if an offenceotherwise warrants a short term of imprisonment, it may be appropriate toreduce the sentence below the imprisonment threshold to a sentence of homedetention or community detention in order to give the guilty plea appropriaterecognition. It may also be appropriate to impose a single sentence instead ofa combination of sentences (for example, supervision instead of supervisionand community work). In cases such as this, a percentage reduction is notpossible. The type and length of sentence that gives appropriate effect to thesliding scale set out above is a matter of judgement.Consistent with that observation, though usually without citing it, judges havesuggested that the two-step approach taken in Reweti is confined to cases that are onthe cusp of home detention and plainly call for a merciful approach.6461 Reweti v R, above n 58, at [24]–[25].62 Peke-Meihana v R, above n 58.63 At [28].64 R (CA217/2018) v R, above n 58, at [26]–[28]; Houkamau v Police, above n 58, at [43];[42] However, the three-step methodology can reduce guilty plea discounts in moreserious cases. This is such a case, as we explain at [57] below, though the effect isminor. Ms Hall, for the intervenor, cited R v R, a sexual violation case in which thesentencing judge adopted a starting point of 12 years then deducted 25 per cent(36 months) for youth and a year for mental health issues and rehabilitative potential,followed by the guilty plea deduction of 25 per cent.65 The resulting sentence was sixyears imprisonment. The two-step methodology would have produced a sentence thatwas shorter by 12 months. On appeal, this Court declined to apply the two-stepmethodology and held that six years was the right sentence in the circumstances.66[43] The cases accordingly confirm that the three-step methodology can reduce thesize of guilty plea discounts for offenders with significant personal mitigating features.Judges have sometimes, but not always, identified the issue and made compensatoryadjustments to the sentence calculation, but that merely confirms the methodologyitself needs adjusting.Should the Court depart from Hessell?[44] Mr Lillico reminded us that this Court ordinarily considers itself bound by itsown decisions and will depart from them only in limited circumstances.67 He pointedout that Hessell was a Full Court decision. But the issue before us concerns thisCourt's supervisory jurisdiction over criminal practice, which by design leads to theCourt dealing with many appeals and invites an evolutionary approach. The Courtmade a similar point in R v Chilton, recognising that a more flexible approach may bewarranted when dealing with "litigation practice", a field in which an intermediateappellate court is likely to know how a given legal doctrine is working in practice.68The Court also noted that a more flexible approach may be necessary in criminalproceedings.69 The methodological issue was not addressed in Hessell. Finally, thePeke-Meihana, above n 58, at [28]; Jones v R, above n 58, at [22]; and Solicitor-General v Kaokao,above n 58, at [32]–[34].65 R v [R] [2018] NZDC 4473.66 R (CA217/2018) v R, above n 58.67 R v Chilton [2006] 2 NZLR 341 (CA) at [83].68 At [96].69 At [103].case for altering the methodology is clear. For these reasons we think there is sufficientcause to depart from this Court's decision in Hessell.A modified methodology for calculating guilty plea discounts[45] The following sentencing methodology replaces the three-step methodologyestablished in this Court's judgment in Hessell at [14].70 It also replaces the Court'ssubsequent restatement in R v Clifford at [60].71[46] A two-step methodology should be used:(a) the first step, following Taueki, calculates the adjusted starting point,incorporating aggravating and mitigating features of the offence;(b) the second step incorporates all aggravating and mitigating factorspersonal to the offender, together with any guilty plea discount, whichshould be calculated as a percentage of the adjusted starting point.72[47] Because the court fixes all second-step uplifts and discounts by reference tothe adjusted starting point under this methodology, it makes no difference to sentencelength if the guilty plea discount is the last step in the sentence calculation.However, the sentencing judge should still quantify a guilty plea discount, for severalreasons: the discount is justified in substantial part by systemic and socialconsiderations distinct from the offender's personal circumstances; the discount mustbe transparent, which aids predictability; and the calculation allows others, includingthe offender and the victim, to identify the sentence that would have been imposed butfor the plea.73 It should be apparent that the discount does not exceed the maximumof 25 per cent of the adjusted starting point.[48] This methodology does not preclude credit for some mitigating factors beingassessed by reference to what would otherwise be the end sentence (that is, the product70 Hessell (CA), above n 1.71 R v Clifford, above n 48.72 We apply the methodology to this appeal at [71] below.73 See Hessell (SC), above n 1, at [73].of step 2), where that is appropriate. For example, credit for time spent onelectronically monitored bail is commonly calculated in that way.[49] As explained at [4] above, guideline judgments such as this one promotetransparency of analysis and principled consistency of outcome, so furtheringobjectives of the Sentencing Act. We repeat that the ultimate question, however, is notwhether an applicable guideline judgment is followed but whether the sentence is ajust one in all the circumstances. When answering it the sentencer should stand backand consider the circumstances of offence and offender against the applicablesentencing purposes, principles and factors.Ms Moses's sentence appealThe facts[50] The police apprehended Ms Moses following a 2017 operation targetingmethamphetamine dealing in Gisborne. Intercepted communications established thatbetween May and December 2017 she received 965g of methamphetamine foron-supply, in a total of 169 transactions.[51] Her practice through most of this period was to receive payment from her owncustomers then place an order with the principal offender, Lucky Campbell. A tableincorporated in the summary of facts records that until October 2017 she usuallyreceived supplies of between one and five grams, with a small number of largertransactions.[52] The quantities received from Mr Campbell increased in October 2017.Between 13 October and 27 November there were eight transactions in each of whichshe received 28g or one ounce. Thereafter there were three transactions between30 November and 7 December in each of which she received 56g or two ounces.This reflected what the summary described as a move to a sales target arrangement inwhich Mr Campbell supplied her with ounce amounts and resupplied her once she hadmet her target. The frequency of the transactions — three of 56g apiece during oneweek — indicates that she was selling significantly more than she had done previouslyand invites the inference that she was supplying other dealers rather than end users.[53] When arrested on 31 January 2018 Ms Moses was found to be in possession of10.8g of methamphetamine and $10,280 in cash. In interview she admitted selling thedrug to pay her bills.The sentencing[54] Ms Moses faced one representative charge of possessing methamphetamine forsupply. Judge Cathcart recorded that the quantity placed her in Fatu band four andobserved that deterrence and denunciation were of primary importance but did notpreclude an allowance for personal circumstances.74 By reference to similar cases headopted a starting point of 11 and a half years imprisonment.75 He appeared to acceptthat she was addicted, although that claim depended on self-report and her claimedusage was modest, but he made no allowance for addiction when setting the startingpoint. Rather, he characterised the offending as commercially motivated.[55] The Judge recognised that at the age of 43, Ms Moses had no relevant priorconvictions, and he was prepared to acknowledge her willingness to undertaketreatment for addiction. He declined an adjournment for a s 27 cultural report,reasoning that it could not make a material difference to sentence given that, followingJarden v R,76 personal circumstances must be subordinated to denunciation anddeterrence.77 Taking all mitigating factors into account he allowed a discount of13 months or nine per cent.78[56] The guilty plea discount allowed was 20 per cent.79 The Crown accepted thatfamily dynamics may have delayed the plea, which also contributed to resolution ofwhat was to have been a lengthy multi-party trial.[57] The sentence calculation followed the three-step methodology. From thestarting point of 11 and a half years (138 months) the Judge deducted 13 months forpersonal mitigating factors, resulting in a provisional sentence of 125 months.74 R v Moses [2019] NZDC 5170 at [2] and [9].75 At [17].76 Jarden v R [2008] NZSC 69, [2008] 3 NZLR 612.77 R v Moses, above n 74, at [28].78 At [29].79 At [31].From that figure he deducted 20 per cent or 25 months for the guilty plea.The calculation resulted in an end sentence of 100 months or 8 years 4 monthsimprisonment, with no minimum period.80 Had the two-step methodology been usedthe end sentence would have been 97.4 months or, when rounded down to the nearestmonth, 8 years 1 month imprisonment: a reduction of three months.[58] We observe that this three-month difference does not in itself warrantinterfering with a sentence of more than eight years imprisonment. It is a smalldifference in a long sentence, and it is unlike a calculation error, which may result inthis Court adjusting the sentence to achieve what the sentencing judge evidentlyintended even if the adjustment is a small one.81 If the Judge applied Hessell correctlythe end sentence was what he considered appropriate after standing back and inquiringwhat was the right outcome. As will be seen, however, we propose to adjust thesentence for other reasons and when doing so we will use the two-step methodology.The appeal[59] For Ms Moses, Ms Thorpe argued that when considered against the Zhangguidelines the starting point was too high, and the Judge also erred by refusing anadjournment for a s 27 report and consequentially giving insufficient weight topersonal circumstances. We have already dealt with the appeal so far as it alsoaddressed the three-step methodology.The starting point[60] The offending falls into Zhang band four: amounts between 501g and 2kg anda starting point between 8 and 16 years imprisonment.82 Ms Thorpe focused on role,arguing that Ms Moses fell into the lower end of the "lesser" role category.Counsel characterised Ms Moses as a low-level street dealer who sold small amountsand depended on Mr Campbell to source the drug. She did not know the full extent ofhis operation and she had no influence upon him. He moved her onto the target system,with "unequivocal expectations that she would sell a global target within certain time80 At [32]–[33].81 Tutakangahau v R [2014] NZCA 279, [2014] 3 NZLR 482.82 Zhang v R, above n 3, at [125].frames". She was not a user when she began but she soon became addicted.Counsel submitted that a starting point of eight and a half to nine years is appropriate.[61] Mr Lillico responded that the starting point remains appropriate under Zhang.He submitted that Ms Moses had a significant role. She was a second-level supplierwhose involvement escalated substantially over time. She was likely moving to awholesaler role; it is unlikely that the large quantities she was handling byOctober 2017 were being sold to end users. She dealt almost every day, and sometimesseveral times a day. She admitted that her motivation was primarily financial.Given that the overall quantity was almost twice the amount needed for entry toband four, it cannot have been an error to take a starting point just below the mid-point.[62] We are not persuaded that the starting point was wrong. The quantity placesMs Moses at the mid-point of band four and although she worked for Mr Campbell itis not accurate to characterise hers as a lesser role. We draw the inference that she hadbecome more than a retail dealer by the time of her arrest. She was not involvedthrough naivety, nor is there evidence that she was engaged through pressure orintimidation. We are prepared to accept that she was addicted, but that does not explainoffending on this scale. Her motivation was financial. She may not have known thefull scale of Mr Campbell's operation, but she was sentenced only for themethamphetamine she dealt herself.Personal mitigating factors: the s 27 report[63] As noted, the allowance for all mitigating factors was 13 months ornine per cent. When declining to adjourn for the s 27 report, the Judge made it clearthat he appreciated the importance of such reports but he took the view that in this casea report would likely make no more than a marginal difference.83[64] Ms Thorpe argued that this was an error. Ms Moses pleaded early, on15 February 2018, and sentencing was scheduled for 20 March. She was entitled topresent information under s 27. The report was in preparation. She argued that thesince-completed report points to significant mitigating factors and a history of social,83 R v Moses, above n 74, at [28].cultural and economic deprivation, all of which may be taken into account followingZhang.84[65] Mr Lillico responded that the Judge appropriately considered personalcircumstances, which were mentioned in the brief pre-sentence report, and the s 27report contains little relevant information. The report does not identify a sufficientlinkage between deprivation and the offending; rather, it details the effect of variousgovernment policies on the Gisborne region without linking them to Ms Moses in away that demonstrates causation between deprivation and offending.[66] We admit the report for purposes of the appeal. It was prepared byMatau Cultural Annotators. It is a thoughtful report. Ms Moses was not an entirelyreliable narrator, which the report attributes to her methamphetamine use, buteventually she underwent extended interviews and disclosed her background.The report explains that her offending followed a bout of depression after the death ofher mother and a sports injury that left her unable to maintain labouring work.It confirms that her motivation was financial and explains she was valuable toMr Campbell because she had no prior drug convictions and could exploit herextensive whānau networks across competing gangs.[67] The report reviews her life circumstances; she is one of a large reconstitutedwhānau and has four children of her own, some of whom have been raised by othersin whāngai adoptions. Although she described her childhood as happy, she alsodisclosed that she was abused as a child within the wider whānau and was not protectedby responsible adults, with consequences for both her whānau bonds and her adultrelationships.[68] The report recounts the Māori rural-to-urban diaspora of the 1960s and 1970sand the consequential loss of social structures, tikanga, culture and language.Existing deprivation was compounded by structural reforms of the 1980s, which"savaged" the Gisborne area and resulted in high unemployment, relative poverty,gang membership and substance abuse. In this setting, whānau are often unsafe,disorganised and ill-resourced. Ms Moses's reconstituted whānau exhibited these84 Zhang v R, above n 3, at [133]–[137].features, and it was in this setting that she was introduced to methamphetaminefollowing the death of her mother and her own inability to work and resultingdepression. The report states that she has sought help for her drug use and notes thatshe is considered a low reoffending risk, although it gives us no cause for optimismabout support from her whānau or community.[69] We accept the report's account of Ms Moses's economically and sociallydeprived background. The strength of any causal connection to her offending isnonetheless difficult to gauge. She had a limited offending history until she beganselling methamphetamine in her forties. She then dealt in substantial quantities tomake money, not because she was addicted or under pressure from associates.It appears that two events led to her offending: the death of her mother, with whomshe had a complex relationship, and her inability to work as a labourer followinginjury.[70] We accept that these events point to a connection between her social andcultural background and her offending. It is sufficiently proximate to mitigateculpability to a degree. The report also points to prospects of rehabilitation which wethink merit recognition. We will increase the discount to 15 per cent (21 months) forthese matters. The guilty plea discount of 20 per cent (28 months) is not in dispute.Sentence calculation[71] The resulting sentence calculation begins with the adjusted starting point,11 and a half years or 138 months. From that we deduct 21 months (15 per cent) forpersonal mitigating factors and 28 months (20 per cent) for the guilty plea. The totalsecond step discount is therefore 49 months and the end sentence is 89 months orseven years, five months imprisonment. Expressed another way, the calculation is138 – (21 + 28) = 89.Result[72] The appeal is allowed. We quash the sentence passed below and substitute asentence of seven years, five months imprisonment.Solicitors:Vicki Thorpe, Gisborne for AppellantCrown Law Office, Wellington for Respondent