Sentencing Guidelines for SA.pdf - SENTENCING GUIDELINES FOR SOUTH

Document Sample
Sentencing Guidelines for SA.pdf - SENTENCING GUIDELINES FOR SOUTH Powered By Docstoc
					                    SENTENCING GUIDELINES FOR
                   SOUTH AFRICA: LESSONS FROM
                                                                 S S TERBLANCHE*
                        Professor of Law, Department of Criminal Law and Procedure, Unisa

In 2000 the South African Law Commission recommended that a sentencing
council be established in South Africa.1 The setting of sentencing guidelines
would be one of the primary functions of this council.2
   The idea of sentencing guidelines is foreign to most South Africans. As a
result there is little clarity as to what is meant by ‘sentencing guidelines’. It
could range from any form of guidance giving to a sentencer, to numerical
sentencing guidelines contained in a grid,as developed in the United States of
America. The object of this contribution is to describe different forms of
sentencing guidelines as applied (or formerly applied) in a variety of foreign
legal systems.3 The article is mostly descriptive,attempting to provide a better
sense of what is meant, or could be meant, in sentencing law when mention
is made of ‘sentencing guidelines’. To put it in a proper South African
perspective it is important to briefly restate the framework within which
sentencing guidelines are proposed for South Africa.

The need for sentencing guidelines
The South African sentencing system requires substantial reform. The Law
Commission found4 that an ideal sentencing system

   * LLD (South Africa). A number of institutions made this research possible. I would like to thank the
Institute of Criminology, University of Cambridge and the Max Planck Institute for Foreign and
International Criminal Law, Freiburg, Germany for the opportunity to spend time there. I would also like to
thank UNISA for the grant that enabled me to travel to Europe, and the NRF for the grant to attend the
Second International Conference on Sentencing and Society in Glasgow (27–29 June 2002).
   1 South African Law Commission Report (Project 82):Sentencing (A New Sentencing Framework) (December

2000) (hereinafter ‘the Report’). Available at [28 October 2003].
Legislation along these lines is on the legislative programme of the Department of Justice and Constitutional
Affairs.More pressing matters seem to have prevented its actual submission to Parliament.Since 17 Jan 2003 the
Law Commission ‘shall be known as the South African Law Reform Commission’ (s 5 of Act 55 of 2002).
   2 See the discussion below.
   3 Since South Africa appears to be taking an African lead with respect to sentencing guidelines, there is,

unfortunately, no material available for an African perspective.
   4 In Report: Sentencing (A New Sentencing Framework) (Project 82) (December 2000) para 1.43 and the

preceding Discussion Paper 91: Sentencing (A New Sentencing Framework) (Project 82) (2000) (hereinafter ‘the
discussion paper’) Part II para 2.5.

SENTENCING GUIDELINES FOR SOUTH AFRICA                                                                    859
    ‘should be seen to promote consistency in sentencing, deal appropriately with concerns that
    particular offences are not being regarded with an appropriate degree of seriousness, allow
    for victim participation and restorative initiatives and, at the same time, produce sentencing
    outcomes that are within the capacity of the State to enforce in the long term’.
    Sentencing guidelines are ultimately aimed at consistency. The first two
aspects that are mentioned in the quotation above focus on consistency. The
lack of consistency in sentencing is a major problem in South Africa, as it is in
other countries where sentencers have a largely unfettered discretion in
imposing sentence.5 It is fairly widely accepted that there are four techniques
that can be used to reduce disparity, namely the enactment of statutory
sentencing principles, various systems of sentencing guidelines, judicial
self-regulation and mandatory minimum sentencing schemes.6
    The Law Commission concluded that the best way of achieving the ideal
system is,first,clearly to state the basic sentencing principles in legislation and,
secondly, to set up a sentencing council to provide sentencing guidelines.7
Before looking at various kinds of sentencing guidelines, let us briefly
consider the proposed ‘basic sentencing principles’.

The basic sentencing principles
The Law Commission’s recommendations regarding the basic sentencing
principles are contained in a Draft Sentencing Framework Bill.8 Overall,
the purpose of sentencing is stated as being to punish offenders for the offence
of which they have been convicted.9 Through this statement the draft
legislation takes a clear stand in favour of retribution as dominant sentencing
   The next general principle is that the seriousness of the crime should
determine the severity of the sentence.The seriousness of a particular offence
has to be established in relation to other offences, and not in a vacuum.10
Proportionality between all offences is, therefore, required. This second
principle is refined by directly connecting it to two specific characteristics,

      5 Such countries include the USA, England, Scotland (cf David M Walker The Scottish Legal System: An

Introduction to the Study of Scots Law 8 ed (2001) 563–7),Australia and the Netherlands.Countries not allowing
for a largely unfettered discretion do not have the same problem — see, e g, the text at note 155 with respect
to the position in Germany. The problem has received little attention in local research, but has been noted in
works such as E M Burchell et al South African Criminal Law and Procedure vol I General Principles of Crminal
Law 2 ed (1983) 80; J K Lund ‘Consistency as a principle in sentencing’ (1983) 7 SACC 3ff; H C Nicholas
‘Consistency and discretion in sentencing in the courts’ in James Midgley et al (eds) Crime and Punishment in
South Africa (1975) 149–56; S S Terblanche The Guide to Sentencing in South Africa (1999) 142–3.
      6 See Andrew Ashworth ‘Structuring sentencing discretion’ and ‘Four techniques for reducing

sentence disparity’in Andrew von Hirsch & Andrew Ashworth (eds) Principled Sentencing:Readings on Theory
and Policy 2 ed (1998) 215–17 and 227.
      7 The Report op cit note 4 para 1.44
      8 The Report op cit note 4 at 99ff.
      9 Section 2. In referring to the ‘offences of which . . . convicted’ the Commission decided against what

is termed ‘real-offence sentencing’ — for a discussion see Michael Tonry & John C Coffee Jr ‘Enforcing
sentencing guidelines: Plea bargaining and review mechanisms’ in Andrew von Hirsch et al (eds) The
Sentencing Commission and its Guidelines (1987) 152–63. Real-offence sentencing is seen by some as the only
way of curbing the effect of plea bargaining on sentencing guidelines. For more on plea bargaining, see the
text at note 184.
    10   Section 3(1). This is in contrast to the view currently mostly favoured, namely that other cases are of
little assistance in establishing an appropriate sentence in the current instance — see the text at note 143.
860                                                   THE SOUTH AFRICAN LAW JOURNAL
namely the degree of harmfulness (or risk of harmfulness) of the offence and the
degree of culpability of the offender.11 Roughly speaking, this means taking into
account the amount of harm involved (or potentially involved) in the
commission of the crime, and the extent to which the offender can be blamed
for this harm.
   Subject to this primary principle of proportionality, the draft legislation
also provides for an ‘optimal combination’of aims towards which the sentence
should strive. These aims are (1) restoring the rights of victims, (2) protecting
society and (3) affording the offender the opportunity of a crime-free life.12
Rather than requiring one of these aims to be selected, one should rather
attempt to discover an optimal combination.
   The sentence proportionate to the seriousness of the offence may be
adjusted ‘to a moderate extent’ by the presence or absence of previous
convictions related to the current offence.13

The guidelines
The primary function of the proposed Sentencing Council will be to draw up
sentencing guidelines. In so doing the Council will be bound by the
above-mentioned basic principles.14
   A guideline is defined in the draft legislation in the following terms:
    ‘A sentencing guideline specifies sentencing options and their severity for a particular
    category . . . or sub-category of offence.’15
A guideline will therefore have to make mention of a particular category or
sub-category of offence, and then specify a sentencing option that should
be imposed for that offence, as well as its quantum. The offences will have
to be graded in terms of seriousness, and ranked accordingly.16 Specific
provision is made for the inclusion of three sentencing options, or any
combination thereof, in the guidelines, namely imprisonment, a fine and a
community penalty.17 A guideline may provide for a range of sentences,
allowing a 30 per cent variation in quantum up or down from the basic
guideline.18 The guidelines may also provide for the suspension of the whole
or part of a sentence.19 There is also provision for the Sentencing Council to
specify specific guidelines for previous convictions, as long as such guidelines
are no more than ‘moderately’ different from the standard guidelines, and as

  11   Section 3(2).
  12   Section 3(3). Points (2) and (3) encompass what would traditionally have been referred to as
deterrence, prevention and rehabilitation, but with a different focus.
   13  Section 3(4). There are a few further basic principles, but they are not of current importance.
   14  Section 5(3) — ‘Sentencing guidelines are determined by applying the sentencing principles in
section 3. . . .’
   15  Section 5(1).Reference to ‘category ...or sub-category of offence’is necessary since our criminal law
is uncodified. It also allows more than one similar type of offence to be included in one offence category.
   16  Section 5(3)(a).
   17  Section 5(2).
   18  Section 5(7). For example, if the basic guideline provides for 30 months’ imprisonment for a specific
category of offence,the guideline may provide for a range of from 21 to 39 months’imprisonment (9 months
being 30% of 30 months).
   19  Section 5(7)(b).
SENTENCING GUIDELINES FOR SOUTH AFRICA                                    861
long as some relationship with the current offence is factored in.20 The draft
legislation leaves little room for the guidelines to be vague.
   In determining the extent of the guidelines account must also be taken of
the capacity of the correctional system, both with respect to prisons and
community corrections.21

On 1 May 1980 Minnesota was the first American state to put into operation a
system of sentencing guidelines.22 The guidelines were the result of the efforts
of the Minnesota Sentencing Commission. At the time it was a revolutionary
new way of sentencing.23
   Even though the Minnesota sentencing guidelines might be reasonably
well-known in South Africa,24 it is useful briefly to look at them again. They
are characterized by a grid, or matrix — a two-dimensional object resembling
a crossword puzzle. An extract from the official grid is represented below in
Figure 1, followed by an explanation of its essential features.

                                                                         Criminal history
           Seriousness level of offence                       0             2            4          6 or
11. Murder, 2nd degree (intentional)                        306           346          386           426
                                                          299-313       339-353      379-393       419-433
8. Aggravated robbery                                        48           68            88           108
                                                            44-52        64-72         84-92       104-112
5. Simple robbery/residential burglary                        18           28           38           48
                                                                                       36-40        46-50
2. Theft crimes ($2 500 or less)                              12           13            17          21

                  FIGURE 1: Extract from Minnesota sentencing grid25

   20  The manner in which previous convictions are crafted into guidelines will always present a major
challenge to any guidelines system.See,for examples of detailed discussion of some of these difficulties,Julian
V Roberts ‘The role of criminal record in the sentencing process’ in Michael Tonry (ed) Crime and Justice: A
Review of Research vol 22 (1997) 303–62 and Andrew Ashworth Sentencing and Criminal Justice 3 ed (2000)
   21  Section 5(3)(b).
   22  Cf Von Hirsch et al op cit note 9 at 177. Some sources erroneously use 1981 as date of
   23  Cf Milton Heumann ‘Plea bargaining: Process and outcome’ (2002) 38 Criminal Law Bulletin 630 at
636. See also David R Lee The Impact of State Sentencing Commissions and Their Guidelines on Prisons and Crime
(2001) (unpublished doctoral thesis, University of Maryland) 9 (‘a courageous and ambitious effort, into
unchartered territory’).
   24  They have been discussed or mentioned in a number of local publications — cf Jan H van Rooyen
‘Doelgerigte straftoemeting: ’n regsvergelykende ondersoek na wyses om straftoemetingsdiskresie te
reguleer’ (1992) 55 THRHR 386–410; D P van der Merwe Sentencing (1991, updated 1998) 5ff.
   25  Minnesota Sentencing Guidelines Commission Minnesota Sentencing Guidelines and Commentary
(2003) 48; effective from 1 Jan 2003 — available at [28
October 2003]. Cf also the new Minnesota Sentencing Guidelines and Commentary (effective from 8 August
2003), available at [28 October 2003].
862                                                     THE SOUTH AFRICAN LAW JOURNAL
   Each row of the grid represents felonies26 of a pre-determined seriousness.
The Minnesota grid contains 11 seriousness levels, of which four have been
included in Figure 1. The columns of the grid represent a pre-determined
representation of the offender’s criminal history,27 in seven levels, from 0 to ‘6
or more’,of which four are included in Figure 1.In the cells where the various
rows and columns intersect, a number plus (sometimes) a range of numbers
can be found, representing the duration of imprisonment that should be
imposed. For example, an offence of level 5 (such as simple robbery) and a
criminal history of level 4 should attract a sentence of 38 months’ imprison-
ment. The italicized numbers in the same cell are the range within which
sentences may be imposed without giving reasons.28 If the cell falls in the
shaded area, a stayed sentence should be imposed rather than direct
imprisonment.29 Obviously, the offence level and the criminal history score
have to be determined before the grid can be employed. This might entail
quite an involved process.
   The Minnesota guidelines are described as ‘presumptive and
prescriptive’.30 They can be departed from in the presence of ‘substantial and
compelling aggravating or mitigating factors’.31 Any departure has to be
motivated, in writing, by the judge imposing the sentence. However, many
circumstances are specifically excluded from consideration. These include
factors such as race, sex, the employment status of the offender and social
factors such as his education or marital status.32 Many of these factors could
influence a particular sentence in South Africa,33 but are excluded from the
Minnesota guidelines in order to curb unfair discrimination among
   Against this background of sentencing grids, it is important briefly to
consider the circumstances that gave rise to this revolutionary way of
determining sentences.

   26   A general definition of a felony is any offence for which imprisonment in excess of 1 year can be
imposed, or that has been declared a felony by statute — cf John C Klotter & Terry D Edwards Criminal Law
5 ed (1998) 16. Therefore, in Minnesota, only the more serious offences are included in the guidelines.
   27   The term ‘criminal history’ is not commonly used in South African law. It is retained in its American
form, as it does not simply reflect the offender’s number of previous convictions.
   28   Von Hirsch et al op cit note 9 at 178.
   29   Minnesota Sentencing Guidelines and Commentary § C, § II. C. 01. The South African equivalent to a
stayed sentence is the deferment of sentencing in terms of s 297(1) of the Criminal Procedure Act, 1977.
   30   Minnesota Sentencing Guidelines and Commentary § C,§ II.C.01.Terance D Miethe & Charles A Moore
‘Socioeconomic disparities under determinate sentencing systems: A comparison of preguideline and
postguideline practices in Minnesota’ (1985) 23 Criminology 337 at 342. ‘Prescriptive’ means they were
intended to change current sentencing practices. A ‘descriptive’ approach would describe the current
sentencing practices — Richard S Frase ‘Sentencing principles in theory and practice’ in Tonry op cit note
20 at 363 at 393.
   31   Minnesota Sentencing Guidelines and Commentary § D, § II. D. 01 (it will be a small number only); cf also
Frase op cit note 30 at 392; Miethe & Moore op cit note 30 at 342–3.
   32   Minnesota Sentencing Guidelines and Commentary § D.1.See also Von Hirsch et al op cit note 9 at 180.
   33   Cf M A Rabie et al Punishment: An Introduction to Principles (1994) 5 ed 310–14, 327 and 338. Our
courts state that all these factors are ‘taken into consideration’ when imposing sentence. It is, however,
indeterminable whether they actually affect the sentences at all. At best the influence would be
negligible — cf, for a more extensive explanation, Terblanche op cit note 5 at 211 and 222–3.
   34   Minnesota Sentencing Guidelines and Commentary § II.D.101 (sentencing should be neutral with respect
to the offender’s sex, race and income levels).
SENTENCING GUIDELINES FOR SOUTH AFRICA                                       863
Historical development of the sentencing-commission idea
The history of sentencing reform in the USA since the early 1970s is well
documented.35 At the time,an indeterminate sentencing scheme was in place.
Sentences were not imposed as a fixed term of imprisonment, but in wide
ranges,such as ‘10 to 25 years’imprisonment’.The prisoner would eventually
be released when the parole board decided that he was sufficiently
rehabilitated. This could either be after a short or long stay in prison or
anything in between. No substantive sentencing law had been developed
through reasoned sentence appeal decisions.36 Disparity was an inescapable
    Then a federal court judge, Marvin Frankel wrote a ‘small volume’ that
would become highly influential, in which he highlighted what he described
as the ‘gross evils and defaults’ in the imposition of sentences.38 As remedy he
    ‘the creation eventually of a detailed chart or calculus to be used . . . by the sentencing judge
    in weighing the many elements that go into the sentence’.39
   This chart or calculus would be developed by a ‘Commission of
Sentencing’,which would be a permanent agency responsible for an effective
program of research and development on sentencing and related fields, and
formulating and enacting the necessary laws and rules.Such a commission was
essential because single judges are not in a position to do the research
necessary for the creation of a rational sentencing system, and because
‘legislative action tends to be sporadic and impassioned’.40
   Around the same time compelling evidence showed that the rehabilitative
ideal of sentencing did not materialize in practice.41 These and a number of
other factors42 contributed to large-scale reforms in sentencing in the USA,
including the creation of sentencing commissions with sentencing guidelines,
much as envisioned by Judge Frankel.43
   Minnesota was followed by other states but,perhaps most significantly,also
by the federal criminal justice system.

   35   See, e g, Michael Tonry Sentencing Matters (1996) 6–13; Franklin E Zimring ‘Sentencing Reform in
the States: Lessons from the 1970’s’ in Michael Tonry & Franklik E Zimring (eds) Reform and Punishment:
Essays on Criminal Sentencing (1983) 101;Frank O Bowman III & Michael Heise ‘Quiet rebellion? Explaining
nearly a decade of declining federal drug sentences’(2001) 86 Iowa LR 1043 at 1050–5.See,in general,Francis
A Allen The Decline of the Rehabilitative Ideal (1981).
   36   Cf Michael Tonry ‘Structuring sentences’ in Michael Tonry & Norval Morris (eds) Crime and Justice:
A Review of Research Vol 10 (1988) 267 at 283.
   37   Cf Franklin E Zimring et al Punishment and Democracy:Three Strikes and You’re Out in California (2001)
   38   Marvin E Frankel Criminal Sentences: Law without Order (1973) vii.
   39   Frankel op cit note 38 at 113.
   40   Frankel op cit note 38 at 119.
   41   Cf the oft-cited work by Robert Martinson ‘What works? — Questions and answers about prison
reform’ (1974) 35 The Public Interest 22.
   42   See Zimring op cit note 35 at 101–21 for an overview of these reasons (the prisoners’ rights
movement; the Attica prison riot; the resultant writings of scholars, both liberal and conservative; politicians
and the law-and-order issue).
   43   Amazingly, Frankel’s proposals are still used as criteria for determining the success of current
guidelines — cf Kevin R Reitz ‘The status of sentencing gguideline reforms in the United States’in Michael
Tonry (ed) Penal Reform in Overcrowded Times (2001) 31ff; Michael Tonry ‘Sentencing commissions and their
guidelines’ in Michael Tonry (ed) Crime and Justice: A Review of Research vol 17 (1993) 137–95.
864                                                     THE SOUTH AFRICAN LAW JOURNAL
The Federal Sentencing Guidelines
The US Sentencing Commission was established by the Federal Sentencing
Reform Act of 1984. The Commission was tasked with drafting guidelines
for the sentencing of federal offenders. Its guidelines became law in 1987,44
and have been unpopular from the very beginning:
    ‘The federal guideline system has been subject to an extraordinarily voluminous amount of
    judicial and academic criticism,45 much of it predictable in view of the defects inherent in
    the original conception. Few would regard the system as a suitable model for future

The structure of the Federal Sentencing Guidelines
The federal sentencing guidelines are contained in the Guidelines Manual.47 It
follows the same basic structure as that of Minnesota.It also uses a matrix,with
the different crime seriousness levels in the horizontal plane,and the criminal
history in the vertical. In contrast to Minnesota’s 11 levels of crime-
seriousness, 43 are used in the federal grid. In order to find the relevant
offence-seriousness level,the sentencing court has to work its way through up
to five different stages. The Sentencing Commission developed its own
measures of criminal history scores, which can be fairly complicated to
determine.48 Thereafter, a further three steps need to be considered, before
the sentencing process is complete. A court is only allowed to impose a
sentence outside the range contained in the relevant cell on the grid in the
rare instances that the court finds
    ‘that there exists an aggravating or mitigating circumstance of a kind, or to a degree, not
    adequately taken into consideration by the Sentencing Commission in formulating the
    guidelines that should result in a sentence different from that described’.49

  44     18 USC § 3551.
  45     See, eg, Tonry op cit note 35 at 72–89 (the ‘most controversial and disliked sentencing reform
initiative in US history’ ); Daniel J Freed ‘Federal sentencing in the wake of guidelines: Unacceptable limits
on the discretion of sentencers’(1992) 101 Yale LJ 1681 at 1685 and the sources referred to in note 10 e g that
the guidelines caused dismay and evasion in the courts); Anthony N Doob ‘The United States Sentencing
Commission Guidelines:If you don’t know where you are going,you may not get there’in Chris Clarkson &
Rod Morgan (eds) The Politics of Sentencing Reform (1995) 199–250 (e g, ‘[n]obody seems to think they have
done a good job’ — at 200);G Thomas Eisele ‘The sentencing guidelines system? No sentencing guidelines?
Yes’(1991) 55(4) Federal Probation 16 at 20 (the guidelines are ‘a dark, sinister,and cynical crime management
program’ that is reminiscent of ‘those systems prevalent in Central and Eastern European countries 150 years
ago’); Kate Stith and José A Cabranes Fear of Judging: Sentencing Guidelines in Federal Courts (1998) 5 (‘Judges,
prosecutors, defense attorneys, and probation officers find themselves operating in a labyrinthine system of
rules devised by a distant and alien administrative agency’); Richard L Nygaard Sentencing: As I see it (2000)
230 (A ‘confounding farrago or technical rules’, containing a ‘myriad of rules, tables, equations, descriptions,
comments, definitions and formulae’); various authors in (1995) 78 Judicature 168–214.
   46    Ashworth op cit note 6 at 234.
   47    United States Sentencing Commission Guidelines Manual (Nov 2002) — hereafter referred to as
USSG. The latest edition became effective on 1 Nov 2002 (available at
TABCON02.htm [28 October 2003]), but was supplemented with effect from 30 Apr 2003 (see [28 October 2003]).
   48    For example, 3 points are awarded for each prior sentence exceeding imprisonment of 1 year plus 1
month (lesser sentences attract fewer points); 2 points are added if the current offence was committed while
the offender was serving another sentence; if the offence was committed less than 2 years after the offender’s
release from prison,having served a sentence of at least 60 days,another 2 points are added — USSG § 4A1.1.
The total number of points are then used to establish the Criminal History Category.
   49    USSG § 5K2. 0.
SENTENCING GUIDELINES FOR SOUTH AFRICA                                865
    The guidelines, or any amendments to them, are prepared by the
Sentencing Commission, and submitted to the United States Congress.
Amendments become law automatically after 180 days, unless a contrary law
is enacted during this time.50

Criticism against the federal guidelines
The ‘extraordinary voluminous amount of . . . criticism’51 against the federal
guidelines essentially amount to the following:
1.     The federal Sentencing Commission did not declare a primary
       rationale for sentencing.52 It decided on an indiscriminate mixture of
       rationales, in contrast to Minnesota, which opted for just deserts.53
2.     The guidelines are complex and unclear.54 They are particularly complex
       when it comes to the sentencing for multiple offences, unfamiliar
       offences, or drug-related offences.55 The complexity makes them
       largely incomprehensible to ordinary citizens and potential offenders.56
3.     The guideline outcomes are harsh.57 This has been seen as a disregard
       for the Sentencing Reform Act’s explicit goal that sentences should be
       ‘sufficient but not greater than necessary to comply with the purposes
       of sentencing’.58 The harshness is most notable in the case of dealing59
       in various drugs, where dealing in crack cocaine attracts sentences ten
       times that of dealing in powder cocaine.60 The extent to which
       guidelines are structured towards severity is also evident from the fact
       that almost all the additional factors that should be taken into account
       are aggravating factors, whilst those that would normally tend to
       mitigate punishment are mostly off-limits.61

  50     28 USC § 994(p). See also Dean J Champion (ed) The US Sentencing Guidelines (1989) 242.
  51     See the quotation at note 46 above.
  52     Cf Ashworth op cit note 6 at 233.
    53   Frase op cit note 30 at 388.As to the importance of making a choice of sentencing aims and policy,see
Andrew von Hirsch ‘Guidance by numbers or words?’ in Martin Wasik & Ken Pease (eds) Sentencing Reform:
Guidance or Guidelines? (1987) 46 at 49–51. In South Africa the draft legislation also declares just deserts
(retribution) as the primary aim — see the text at note 9 above.
    54   Editorial ‘Time to re-examine the US sentencing guidelines and mandatory penalties’ (2000) 83
Judicature 176 (a ‘bewildering array of arithmetical adjustments, guidelines,and policy statements ...’).One of
the commissioners of the original commission accepted that a ‘comprehensive system that will reduce
disparity appears to require undesirable complexity’,yet found the guidelines ‘overwhelmingly’complex and
at the same time not really reducing discretion — Paul H Robinson ‘Dissent from the United States sentencing
commission’s proposed guidelines’(1986) 77 Journal of Criminal Law and Criminology 1112 at 1121–2.See also
Albert W Alschuler ‘The selling of the sentencing guidelines’ in Champion op cit note 50 at 61.
    55   The USSG contains extensive provisions on drug offences, totalling some 40 pages in USSG § 2D1. 1
to § 2D3. 2.
    56   Nygaard op cit note 45 at 230.
    57   The harshness and rigidity are two features of the guidelines most disliked by judges — Tonry op cit
note 35 at 13.
    58   Editorial op cit note 54 at 176. See also Sara Manaugh Judges and Sentencing (2002) 102.
    59   In South African criminal-law terms.
    60   In terms of USSG § 2D1. 1(c) the base offence level for 150 kg or more of cocaine is the same as for
1,5 kg or more of ‘Cocaine base’ (i e crack cocaine), namely level 38. Even for a first offender that level
amounts to a range of 235 to 293 months’ (i e some 20–24 years’) imprisonment. See also Bowman & Heise
op cit note 35 at 1131 (sentences for drug offences are very long in comparison to other crimes, and to those
considered appropriate pre-guidelines); Peter H Rossi et al ‘Just punishments: Guidelines sentences and
normative consensus’ (1997) 13 Journal of Quantitative Criminology 267 at 287.
    61   USSG Part H.
866                                                    THE SOUTH AFRICAN LAW JOURNAL
4.        The imbalances in sentences (such as in the case of powder and crack
          cocaine) shows that the system is not rational. Many other examples
          exist, or have existed.62
5.        The guidelines are too rigid, leading to much injustice. It does not
          provide for sufficient discretion to prevent such injustices.63

The Sentencing Commission: current status
All the criticism has not left the Sentencing Commission unscathed.For some
time there were no commissioners, and no work was done.64 In 1999 a new
set of commissioners was appointed, with the chairperson promising to do
something about the criticism.65 Currently most of the criticism remains. It is
unlikely that Congress will allow fundamental changes to be made in the
foreseeable future. In fact, through the so-called PROTECT Act (Pub. L.
108-21, which became effective on 30 April 2003) Congress has recently
directly amended the Sentencing Guidelines Manual, inter alia creating strict
limits on downward departures in sex offence cases.

Sentencing guidelines in the states
The second state after Minnesota to employ sentencing guidelines,created by
a sentencing commission,was Pennsylvania.66 Since then,sentencing commis-
sions have, to a certain extent, become the trend in guiding sentencing. At
times up to 17 states had sentencing guidelines developed by such commissions;67
25 states had at one time or another created sentencing commissions.68 More
such systems are currently in the process of development.69
    The fact that there are sentencing guidelines in several states does not mean
that these systems are all basically the same. On the contrary, the guidelines
in the different states often, have very different characteristics.70 Compliance

     62 Cf for some of these examples, Robinson op cit note 54 at 1112–25.
     63 Manaugh op cit note 58 at 102.
     64 Cf Editorial op cit note 54 at 176. After Richard P Conaboy resigned as chairman on 31 Oct 1998,
almost a year before his term expired, the Commission was without commissioners for more than a
year — Anon ‘Briefs’ (2000) 83 Judicature at 210.
   65   Cf Diana E Murphy ‘Inside the United States sentencing commission: Federal sentencing policy in
2001 and beyond’ (2002) 87 Iowa LR 359–99.
   66   In 1982. Cf J Kramer ‘The Evolution of Pennsylvania’s sentencing guidelines’ in Michael Tonry &
Kathleen Hatlestad (eds) Sentencing Reform in Overcrowded Times (1997) 57.
   67   Richard S Frase ‘Sentencing Guidelines are ‘‘alive and well’’ in the United States’ in Tonry &
Hatlestad op cit note 66 at 12.
   68   Tonry op cit note 35 at 10.
   69   Reitz op cit note 43 at 34 and 39;Manaugh op cit note 58 at 96.The establishment of an ‘agency’very
similar to a sentencing commission for every state has been recommended by the American Bar
Association — see Criminal Justice Standards for Sentencing Alternatives and Procedures 3 ed (1994). There has
been considerable incentives from Congress for states to establish guidelines — Tonry op cit note 35 at 6 and
pers com [9 Jan 2003].
   70   Leonard Orland & Kevin R Reitz ‘Epilogue: A gathering of state sentencing commissions’ (1993) 64
Univ of Colorado LR 837–45; Richard S Frase ‘Sentencing Guidelines in Minnesota and Other American
States: A Progress Report’ in Clarkson & Morgan op cit note 45 at 172–3.
SENTENCING GUIDELINES FOR SOUTH AFRICA                                        867
is purely voluntary in some states.71 Not all of them consist of a matrix,72
while some use different matrices for different kinds of offences.73 Not all the
systems are equally effective, and in some states the guidelines that were
developed have been abolished.74
   Many of the states’ guidelines systems are quite hard to research from a
distance, since little has been written in legal and criminological literature.75
Minnesota is one exception.Its guidelines have been subjected to a substantial
number of assessments, in which it was invariably found that they had
achieved most of their set purposes.76 It is clear that Minnesota’s guidelines
have altered the way in which sentences are imposed. They have reduced
disparity and have done so without over-burdening the correctional system.77
The state’s Supreme Court has developed a jurisprudence of permissible and
impermissible departures from the guidelines.Some criticism remains,such as
that the guidelines are limited to felonies,and that there are no guidelines with
respect to offences not leading to imprisonment.78 However, it is a fact that
the Minnesota Sentencing Commission and its guidelines are held in high
regard by most commentators.79 It remains a model to be taken seriously by
anybody and any jurisdiction considering sentencing grids as a possible
sentencing model.80

Defects of sentencing guidelines in general
The gridlike sentencing guidelines are,conceptually,not above criticism. The
first point is that they tend to stress only the two factors common to most of
them, namely the seriousness of the offence and the offender’s criminal
record, whereas sentences might reasonably be affected by other factors as
well.81 Furthermore, many commentators simply do not like the appearance
of a sentencing matrix and argue that it is a too mechanical way of imposing

   71    Eg Delaware, Utah, Wisconsin — Frase op cit note 70 at 173. Around 1983 there were guidelines
acitivities in 35 states, but evaluations on voluntary guidelines concluded that they had little or no effect on
sentencing decisions — Tonry op cit note 36 at 279–82; Tonry & Hatlested op cit note 66 at 19. See also
Miethe & Moore op cit note 30 at 360.
   72    Eg Virginia — Lee op cit note 23 at 41–2.
   73    Eg Kansas uses a different matrix for drug offences than for others, Utah has a separate matric for sex
offences and Michigan uses several matrices — Lee op cit note 23 at 35–41.
   74    Eg Florida — Reitz op cit note 43 at 33.
   75    Cf Tonry op cit note 36 at 272 (referring to the sparseness of evaluating literature).Some information
on Ohio, Pennsylvania and North Carolina can be found in Tonry & Hatlestad op cit note 66 at 69ff, and
some on Delaware, Utah and Kansas in Tonry & Hatlestad op cit note 66 at 88ff.
   76    The Minnesota Sentencing Commission did much of the evaluation itself — cf Kay A Knapp
‘Implementation of the Minnesota guidelines: Can the innovative spirit be preserved?’in Von Hirsch et al op
cit note 9 at 127; Kay A Knapp ‘Impact of the Minnesota sentencing guidelines on sentencing practices’
(1982) 5 Hamline LR 237–70.But these have been corroborated by independent evaluation as well — Tonry
& Coffee op cit note 9 at 151; Miethe & Moore op cit note 30 at 337–63; Frase op cit note 70 at 196.
   77    Cf Lee op cit note 23 at 17.
   78    Tonry op cit note 36 at 310.
   79    Cf Ashworth op cit note 6 at 232; Zimring et al op cit note 37 at 213; Miethe & Moore op cit note 30
at 344.
   80    Frase op cit note 70 at 198.
   81    Cf Tonry op cit note 35 at 11 (including ‘ethically relevent factors’such as the effect of the sentence on
the accused, the accused’s mental health or drug dependence, a severely deprived background, or
victimization by sexual abuse);Stith & Cabranes op cit note 45 at 5 (individual characteristics of offenders are
generally ignored).
868                                                     THE SOUTH AFRICAN LAW JOURNAL
sentence.82 Sentencing matrix’s are also prone to abuse if extreme care is not
taken in putting them together. This happened in Minnesota, where prosecu-
tors started obtaining multiple-count convictions, since that automatically
produced high criminal-history ratings.83 Fortunately, if the proper
evaluation mechanisms are in place such practices can be picked up and
    Another claim that has attracted a lot of attention, is that the guidelines
have, through plea bargaining, simply shifted the sentencing discretion from
the court to the prosecutor.84 Whether this is a major problem is open to
argument. It is an almost inevitable outcome when the court’s sentence is
known in advance with substantial certainty. Nevertheless, any sentencing
commission should be aware of this problem and should attempt to minimise
its influence, while also being aware that effective measures to counter it has
not yet been devised.

Other experiences with matrices
For the sake of completeness mention should be made of the matrix that was
enacted in Western Australia towards the end of 2000.85 As a result of a new
government coming into power in February 2001 the scheme was, however,
never proclaimed.86

Whereas the pace of sentencing reform in the United States has slackened
considerably,it is continuing unabated in England and Wales.The most recent
development involve the proposed creation of a Sentencing Guidelines
Council, tasked with the framing of sentencing guidelines. This might appear
to involve a step as radical as the grids from across the Atlantic.Before jumping
to this conclusion, however,it is necessary briefly to chart the development of
this council in English law.

   82  Cf Tonry op cit note 35 at 20 (‘grids affect the way people think . . . about sentencing . . . [They
contain] vast amounts of information’) and 98 (referring to the ‘sentencing machine’ problem).
   83  See Knapp op cit note 76 at 129.
   84  See, eg, Heumann op cit note 23 at 630–41 (the sentencing guidelines have moved the discretion to
the prosecutor, to the backroom); Clarkson & Morgan op cit note 45 at 10. For a less critical view, see Ellen
Hochstedler Steury ‘Prosecutorial and Judicial Discretion’ in Champion op cit note 50 at 93–110.
   85  For a general discussion of this legislation, see Neil Morgan ‘Accountability, transparency and justice:
Do we need a sentencing matrix?’ (1999) 28 Univ of Western Australia LR 259.
   86  Neil Morgan ‘Mandatory sentencing, sentencing grids and guideline judgments in Australia’ paper
delivered at the Sentencing and Society Second International Conference (27–29 June 2002), Glasgow. The
complex nature of this legislation,the fact that it was never introduced and therefore not evaluated,and that it
contains further aspects foreign to South African law would make a further discussion of this project
SENTENCING GUIDELINES FOR SOUTH AFRICA                                        869
Guideline judgments in English law
Sentencing disparity has long been considered a problem in English law.Partly
due to this concern Lawton LJ pioneered guideline judgments in the 1970s.87
This practice has been followed by Lord Chief Justices ever since.88
   A guideline judgment is a judgment issued by the Court of Appeal when it
suggests guidelines with respect to a particular category of offence.89 A typical
guideline judgment would begin by stating the current pattern of offending
and sentencing.90 It would consider the basic framework applicable to
sentencing in general and regarding the offence in particular. It would then
provide a fairly exhaustive list of aggravating and mitigating factors associated
with the offence. And then, in many instances, it would actually provide
starting points for sentences for certain sub-categories of the particular
   Many judgments do not go as far as actually suggesting a starting point for
sentences, especially when the variations of that crime are so great that it
would be inappropriate to do so.92 In such cases the court might for example
merely state, as was done in R v Brewster,93 that the degree of seriousness
attaching to burglary is not such that a non-custodial sentence could never be
justified or,as was done in R v Boswell,94 that punishment for causing death by
reckless driving95 should in many cases ‘involve immediate loss of liberty’.
   The Court of Appeal has been careful to stress that the guideline
judgments do not deprive the sentencer of his or her sentence discretion:
    ‘But the sentencer retains his discretion within the guidelines,or even to depart from them if
    the particular circumstances of the case justify departure.’96

   87   In cases such as R v Willis (1974) 60 Cr App R 146 (for buggery);R v Taylor,Roberts and Simons (1977)
64 Cr App R 182 (for unlawful sexual intercourse).
   88   Ashworth op cit note 20 at 30; Martin Wasik Emmins on Sentencing 4 ed (2001) 373. Examples include
R v Barrick (1985) 7 Cr App R (S) 142 (breach of trust theft); R v Boswell (1984) 6 Cr App R (S) 257 (causing
death by bad driving);R v Billam (1986) 82 Cr App R 347 (rape); A-G’s Reference (No 1 of 1989) (1989) 11 Cr
App R (S) 409 (other sexual offences); R v Aramah (1982) 4 Cr App R (S) 407 (drug offences); R v Brewster
[1998] 1 Cr App R (S) 181 (domestic burglary).
   89   Cf R v de Havilland (1983) 5 Cr App R (S) 109. See also Ashworth op cit note 20 at 30 (a judgment
containing the general parameters for several variations of a certain type of offence,the main aggravating and
mitigating factors, and (often) suggesting an appropriate starting point or range of sentences); J J Spigelman
‘Sentencing guideline judgments’ (1999) Current Issues in Criminal Justice 5 at 11. The vast majority of Court
of Appeal judgments are not guideline judgments, but only concerned with the facts and circumstances of
that particular case: ‘It is therefore necessary to read Court of Appeal decisions with care in order to
distinguish those which are intended to provide general guidelines from those which are ‘‘no more than
examples’’ ’ — Nigel Walker & Nicola Radfield Sentencing: Theory, Law and Practice 2 ed (1996) 13.
   90   To this end the Home Office provides the necessary statistics — cf R v Brewster [1998] 1 Cr App R (S)
181 at 184.
   91   This has been described as the main function of guideline judgments — Wasik op cit note 88 at 373.
For example,in R v Boswell (1984) 6 Cr App R (S) 257 at 261,drivers causing death while racing on highways,
or driving recklessly after taking alcohol or drugs should expect two years’imprisonment or more,and being
disqualified from driving for a long time.
   92   Spigelman op cit note 89 at 11.
   93   [1998] 1 Cr App R (S) 181 at 187.
   94   [1984] 6 Cr App R (S) 257 at 259.
   95   At the time a contravention of s 1 of the Road Traffic Act 1972, as substituted by s 50(1) of the
Criminal Law Act 1977.
   96   R v de Havilland [1983] 5 Cr App R (S) 109 at 114.
870                                                  THE SOUTH AFRICAN LAW JOURNAL
   Thus presiding officers must still tailor their sentences to the facts of the
particular case.97 Despite technically being obiter dicta,98 the influence of
guideline judgments has been substantial and these guideline judgments are
considered as being quite successful in reducing disparity for the offences they
deal with. The most prominent of authors on sentencing in England,
Professor Andrew Ashworth, comments as follows:
     ‘Guideline judgments are an innovation of which the senior judiciary can rightly be proud.
     They show how guidance can be fashioned, in a judge-friendly way, based on experience,
     and shaping discretion without constraining it too tightly. It might also be claimed that
     guideline judgments are capable of changing judicial sentencing practices ...although where
     there is a strong judicial culture the guideline judgment may be less than fully successful in
     altering sentencing practice.’99

The deficiencies of guideline judgments
1.   Guideline judgments are delivered sporadically and not as part of an
     overall strategy to reduce disparities of all sentences. They may succeed
     in improving consistency with respect to the specific category of
     offence to which they pertain, but they do not consider its relationship
     with other offence categories.100
2.   Although there are exceptions, most judges are not particularly expert
     regarding the intricasies of sentencing theory and policy.As Von Hirsch
             ‘Common-sense beliefs about deterrence or treatment or dangerousness are not
             necessarily correct; questions about fair distribution of penalties do not necessarily
             resolve themselves easily.’101
3.        Guideline judgments invariably only cover the really serious cases.102
          Offences that are typically heard at magistrates’ courts level are hardly
          ever covered in guideline judgments at all.103 This does not leave
          magistrates completely in the dark. Sentencing guidelines have been in
          use in the magistrates’ courts for some time since they have originally
          been developed for traffic offences. The latest version of the Magistrates’
          Court Sentencing Guidelines, in which ‘important advice and guidance
          on sentencing matters’ is available to magistrates,104 was published
          in 2000.105

     97 Ashworth op cit note 20 at 31.
     98 Ashworth op cit note 6 at 229.
     99 Andrew Ashworth ‘The Decline of English Sentencing and Other Stories’ in Michael Tonry &
Richard S Frase (eds) Sentencing and Sanctions in Western Countries (2001) 62 at 74.
   100  Ashworth op cit note 6 at 228–9.
   101  Von Hirsch op cit note 53 at 58.
   102  Only cases reaching the Court of Appeal. Appeals from magistrates’ courts are mostly heard in
Crown Courts, and are mostly not reported.
   103  Wasik op cit note 88 at 299.
   104  Wasik op cit note 88 at 300. See also Ashworth op cit note 99 at 74–5 (the guidelines contain the
maximum penalty, relevant statutory and other factors, refer to some common factors affecting seriousness,
and propose a starting point for the sentence; they are as significant as the guideline judgments).
   105  See [10 Dec 2002]. New
sentencing guidelines (the 6th ed) will be implemented on 1 January 2004.
SENTENCING GUIDELINES FOR SOUTH AFRICA                            871
  In an attempt to counter some of these deficiencies, the Sentencing
Advisory Panel was introduced.

The Sentencing Advisory Panel
The Sentencing Advisory Panel began its work on 1 July 1999, under the
Crime and Disorder Act, 1998. This legislation places a statutory duty on
the Court of Appeal to consider producing new sentencing guidelines and to
revise existing ones whenever there is an appeal against a sentence.106 If the
Court decides to issue guidelines it has to advise the Panel107 of this
intention.108 The Panel then has to consult with certain predetermined
organizations,109 and eventually formulate its own views which are then
communicated to the Court.110 Thus, the Panel is
    ‘intended to operate as a source of guidance, ideally assisting the Court by researching
    the subject more thoroughly and bringing a wider range of views to it than the Court is able
    to do but . . . the powers of the Court of Appeal in respect of guideline judgments are
    little altered’.111
   The Panel is also entitled to propose sentencing guidelines of its own
accord.112 The Panel’s views constitute one of the factors that the Court has to
consider before framing the guidelines. Others include current sentencing
practices, the cost and effectiveness of the various sentences, as well as less
tangible matters such as the need for consistency and the need to promote
public confidence in the system.113
   The Panel’s advises are basically set in the same narrative format as that of
the guideline judgments.It is typically accompanied by an in-depth discussion
of all matters related to sentencing of that particular offence.114
   The Court of Appeal has adopted the Panel’s advice in a number of
cases.115 In one of the latest instance this involved a review of the guidelines on

Further developments
Major concern about crime in England keeps the effectiveness of sentencing
in the newspapers and,as a result,it is a political topic.The latest governmental

  106    Section 80(1) and (2).
  107    The Panel currently consists of a number of academic experts on sentencing, members of the
judiciary, and representatives from related professions such as magistrates, the probation service and the
   108   Section 81(2).
   109   For a list, see Sentencing Advisory Panel Annual Report (1 April 2001 — 31 March 2002) (2002)
at 25.
   110   Section 81(4).
   111   Ashworth op cit note 20 at 48.
   112   Section 81(3).
   113   Section 80(3).
   114   For the latest advice see [29 October 2003].
   115   Cf R v Kelly and Donelly [2001] Crim LR 411; [2001] 2 Cr App R (S) 73 (racially aggravated
offences);R v Mashaollahi [2001] 1 Cr App R (S) 330 (dealing in opium);R v Webbe [2002] 1 Cr App R (S) 82
(handling in stolen goods) and R v Nelson [2002] 1 Cr App R (S) 565 (extended sentences).
   116   R v Millberry [2003] 1 WLR 546 (CA); [2003] 2 All ER 939; [2003] 1 Cr App R 25.
872                                                    THE SOUTH AFRICAN LAW JOURNAL
report on sentencing, the so-called Halliday report,117 was published towards
the end of 2001. One of its specific proposals was the institution of a kind of a
sentencing commission. These proposals have been given effect to in the
latest Criminal Justice Bill.118 Section 149(1) declares that there will be a
Sentencing Guidelines Council, chaired by the Lord Chief Justice. The
composition of the Council is to be determined by the Lord Chancellor and
the Secretary of State.119 Members will be drawn from the judiciary120 and
from others with experience in the criminal justice system.121 The Sentencing
Advisory Panel will remain in place.122 Essentially, the proposed relationship
between the Panel and the Council has been drafted in the same language as
the current relationship between the Panel and the Court of Appeal.123 Once
guidelines have been set, courts will have to ‘have regard’ to any applicable
guideline.124 In its White Paper on the topic125 the government explained
that the Council will be responsible for setting guidelines for the full range of
criminal offences. How this will be done is not explained in the draft
legislation126 and will clearly have to be decided by the Council itself. Also,
the judge or magistrate
    ‘will continue to make his or her own decision as to sentence,but will be required to operate
    within the Council’s guidelines or explain why they do not apply to the case in question’.127
   On the face of these materials it appears that the main change from the
current situation is going to be that the Sentencing Guidelines Council will
be tasked with setting up guidelines for all offences, and that the Secretary of
State will be given a specific new role in this respect. Time will tell whether
these changes will have a noticeable effect on sentencing practices. However,
what is clear is that the Sentencing Guidelines Council will to build on
English traditions, and will be a body quite distinct from the American
sentencing commissions.

   117   Home Office The Halliday Report (Making Punishments Work: Review of the Sentencing Framework for
England and Wales) (2001) — available at [29 October 2003].
See,for some critical discussion, Estella Baker & C M V Clarkson ‘Making punishments work:An evaluation
of the Halliday Report on sentencing in England and Wales’ [2002] Criminal LR 81.
   118   As amended in Committee on 15th October 2003 available at
ld200203/ldbills/111/2003111.htm [29 October 2003].It is likely to become the Criminal Justice Act 2003.
   119   Section 149(1)(b) and (c).
   120   Including ‘a District Judge (Magistrates’ Courts)’ and lay justices — s 149(2), (3) and (9).
   121   The details are set out in s 149(4)–(6).
   122   Section 151. In terms of s 151(1) it is constituted by the Lord Chancellor after consultation with the
Secretary of State and the Lord Chief Justice.
   123   In s 152(5) and s 153. As to the content, see the text at note 113 above.
   124   Section 154(1)(a).
   125   Home Office Justice for All (2002) 89–90.
   126   Apart from requiring the Council to frame sentencing guidelines when such guidelines are
‘proposed’ by the Sentencing Advisary Panel or the Secretary of State — s 152(3).
   127   Op cit note 125 at 90.
SENTENCING GUIDELINES FOR SOUTH AFRICA                                                                     873
Recent developments in Australia
The guideline judgments of the English Court of Appeal have,at times,spilled
over into other commonwealth jurisdictions, such as Australia, New Zealand,
Hong Kong and Canada.128
    In Australia guideline judgments have mainly been handed down in New
South Wales,129 under the initiative of Spigelman CJ.130 The first of his
judgments specifically referred to as a guideline judgment was R v Jurisic.131 It
was generally well received and was subsequently bolstered by legislation
allowing the Attorney-General (amongst others) to request a guideline
judgment from the Court of Criminal Appeal.132
    The guideline judgments prompted Zdenkowski to make this forecast:
    ‘Judicial sentencing guidelines represent a pre-emptive strike against sentencing grids/
    matrices and such guidelines have a rosy future because they are an incremental development
    and have strong judicial and professional support.’133
   Of course he was not to know that the Australian High Court would cast a
cloud over guideline judgments in Wong v The Queen,Leung v The Queen.134 It
found the guidelines in one of the guideline judgments, R v Wong,135 to be
invalid, and incompatible with federal legislation. However, due to different
considerations motivating the judgments of the nine justices, this judgment
has been described as unsatisfactory ‘because of the difficulty of extracting
clear guidance from it’.136 What does appear certain is that typical guideline
judgments have little future in the context of federal laws.137

  128    See Wong v The Queen, Leung v The Queen [2001] HCA 64 para 9.
  129    For an overview of the rise and fall of interest in guideline judgments in other parts of Australia, see
Kate Warner ‘The Role of Guideline Judgments in the Law and Order Debate in Australia’paper delivered at
Sentencing and Society: Second International Conference (27–29 June 2002) Glasgow, Scotland 10–15. For
example, in Western Australia legislation was introduced in 1995 to authorize the referral of cases to the
Court of Criminal Appeal specifically to formulate sentencing guidelines — s 143 of the Sentencing Act
1995 (WA). However, every time the court was requested to issue guidelines it declined — George
Zdenkowski ‘Sentencing Trends: Past, Present and Prospective’ in Duncan Chappell & Paul Wilson (eds)
Crime and Criminal Justice System in Australia: 2000 and Beyond (2000) 161 at 176. Yet, according to Neil
Morgan and Belinda Murray ‘What’s in a name? Guideline judgments in Australia’ (1999) 23 Criminal Law
Journal 90 at 96–106 this court actually went further in structuring sentencing discretion than the New South
Wales court did,but it has so far refused to elevate its judgments into the more formal guideline judgments.
   130   It was a specific aim of his guideline judgments to reinforce public confidence in the sentencing
process,to increase awareness of appellate guidance and to address inconsistency in sentencing — Warner op
cit note 129 at 3. The Chief Justice played an unusually active role in promoting these judgments publicly,
including writing a newspaper article (Daily Telegraph, 13 October 1998) — Zdenkowski op cit note 129 at
177. See also J J Spigelman ‘Sentencing guideline judgments’ (1999) 73 Australian LJ 876 and Spigelman op
cit note 89 at 5–16.
   131   (1998) 45 NSWLR 209 (1998) 101 A Crim R 259). See Morgan & Murray op cit note 129 at 91–6
for a general discussion. Other guideline judgments include R v Henry (1999) 46 NSWLR 346; Re Attorney
General’s Application [No 1] (1999) 48 NSWLR 327; R v Wong (1999) 48 NSWLR 340 and R v Thomson
(2000) 49 NSWLR 383.
   132   Criminal Procedure Act 1986, as amended by the Criminal Procedure (Sentencing Guidelines) Act
1998 and the Crimes (Sentencing Procedure) Act 1999.
   133   Op cit note 129 at 180. See also Morgan op cit note 86 at 27.
   134   [2001] HCA 64.
   135   (1999) 48 NSWLR 340.
   136   Warner op cit note 129 at 6. See also Morgan op cit note 86 at 27.
   137   See Kate Warner ‘Case and Comment’ (2002) 26 Criminal LJ 59 at 60.
874                                                 THE SOUTH AFRICAN LAW JOURNAL
    However, this is not the end of guideline judgments in Australia. In R v
Place138 the South Australian Supreme Court had little difficulty in
distinguishing its ‘setting of standards’ with respect to drug offences and
armed robbery from guideline judgments, finding that the court may identify
a range of penalties for the ordinary case.139 Similarly, in R v Whyte,140 the
New South Wales Court of Criminal Appeal found that Wong did not require
it to overrule guideline judgments such as Jurisic. Furthermore, numerical
guidelines have a role to play in achieving equality of justice where there is
tension between the principle of individualized justice and the principle of

South African experience with guideline judgments
It might well be argued that South African law has no experience with
guideline judgments.Many years ago it was stated in R v Mapumulo142 that the
infliction of punishment is ‘pre-eminently a matter for the discretion of
the trial Court’.143 This has often been interpreted as indicating a leave-the-
court-alone approach, except that applicable statutory limitations have to
be heeded. The approach is embodied in another oft-cited dictum, from
R v S,144 that ‘rules that bind the Court’ are out of place in the field of
sentencing. As a result our courts have often held it to be an ‘idle exercise’ to
try to match the facts of one case with that of another.145
    Despite a generally strong adherence to these views, some judgments have
at times been quite strict about the sentences that can and cannot be imposed
by the magistrates’ courts. One example is driving under the influence of
liquor.146 In S v Serabo and five similar cases147 the Eastern Cape High Court
recently dealt with a decision by the magistrates of East London to increase
sentences for this offence. In restating certain established principles the court
began as follows:
    ‘It is desirable for sentences for these offences to be consistent so that similar sentences are
    imposed for similar offences. . . .’148
   The court then considered previous sentencing judgments for ‘driving
under the influence’, before finding that the concerted effort to impose
higher, but still consistent sentences, is permitted, but should not lead to
arbitrary sentencing — the ‘norm’ should not be seen as more than a point
of departure.149 A ‘norm’ fine within the range of R4 000 to R6 000 is

  138   [2002] SASC 101.
  139   Paras [26] and [31].
  140   [2002] NSWCCA 343 (20 August 2002).
  141   Guideline judgments have also been bolstered by the legislation referred to in note 132 above.
  142   1920 AD 56 at 57.
  143   This has been followed in a long line of decisions — see Terblanche op cit note 5 at 133n89.
  144   1958 (3) SA 102 (A) at 104D.
  145   Cf S v Fraser 1987 (2) SA 859 (A) at 863C–D and the other cases referred to in Terblanche op cit
note 5 at 129.
  146   The contravention of s 65(1) and (2) of the National Road Traffic Act 93 of 1996.
  147   2002 (1) SACR 391 (E).
  148   At 394.
  149   At 397e.
SENTENCING GUIDELINES FOR SOUTH AFRICA                                      875
considered not to be ‘excessive or unreasonable’150 for this offence, while the
alternative imprisonment ‘should not be more than about eight months, and
in most cases it will be considerably less’.151
   There is no doubt that Serabo is a guideline judgment. It is true that it has
not been passed by the highest criminal appeal court, and is not binding
nationally. Such a judgment might not have everybody’s approval, but its role
in attempting to achieve justice where the principles of individualization and
consistency are at odds with one another152 cannot be doubted.

The Netherlands is one of only a few countries on the European continent
where there is concern over sentencing disparity.153 In most of the others
there is little discussion of sentencing guidelines or basic reform of the
sentencing systems. For example, in Germany sentences have to be within
the broad range provided by the Criminal Code for each offence.154 However,
within this range there are ‘tariffs’ understood by all role players in the
criminal justice system:
    ‘Trial courts are guided . . . by traditional standards and conventions that differ locally. The
    resulting moderate inequality of sentences is regarded with astonishing equanimity by
    practitioners and theorists.’155
   The position is different in the Netherlands, however, where the
sentencing discretion of the judiciary is very wide;156 much wider,in fact,than
in any of the surrounding countries.157

  150    At 399a.
  151    At 400c.
  152    See the text at note 141 above.
   153   Cf P C Vegter ‘Herziening van het sanctiestelsel: Verslag van de jaarvergadering van de Nederlandse
Juristen-Verening’(2002) 77 Nederlands Juristenblad 1288 at 1291;Peter J P Tak Essays on Dutch Criminal Policy
(2002) 125; G J M Corstens Het Nederlands strafprocesrecht 2 ed (1995) 666. A special edition of Trema
(Tijdschrift voor de rechterlijke macht) was dedicated to a discussion of consistency in sentencing — see (1992)
Trema 77–135, in particular A F M Brenninkmeijer ‘Gelijkheid van strafen’ (1992) Trema 77–84. It was the
cause of much discussion. The desire for greater consistency has been expressed for a long time, and many
suggestions have been made as to how it could be achieved — cf J Silvis in He G van de Bunt et al (eds)
Richtlijnen van het openbaar ministerie (1993) 98 at 98–100; C Fijnaut ‘Straftoemeting: een vergeten probleem’
(1989) 19 Delikt en Delinkwent 925–7.
   154   For example, theft is defined in § 242 of the German Criminal Code (Strafgesetzbuch) and declared
punishable with imprisonment of up to 5 years, or a fine.
   155   Thomas Weigend ‘Sentencing and punishment in Germany’ in Tonry & Frase op cit note 99 at
205–6 (source references omitted).See also Hans-Jörg Albrecht ‘Sentencing and punishment in Germany’in
Tonry op cit note 43 at 145 (the current state of the sentencing process is considered to be satisfactory). The
same applies to countries such as France, Spain, Italy and Sweden.
   156   Cf, among many possible references, C P M Cleiren & J F Nijboer (eds) Strafvordering: Tekst &
Commentaar 3 ed (1999) 868; J de Hullu ‘De strafmotivering’ in G Knigge (ed) Leerstukken van Strafprocesrecht
5 ed (2001) 229 (‘de enorme vrijheid’ of Dutch judges); Peter J Tak ‘Sentencing and punishment in the
Netherlands’ in Tonry & Frase op cit note 155 at 174–5.
   157   See P J P Tak ‘Opportuniteitsbeginsel en minimumstraf ’(2002) 32 Delikt en Delinkwent 356–66 for a
comparison of the sentencing discretion in the Netherlands with other European countries, notably
876                                                       THE SOUTH AFRICAN LAW JOURNAL
Role of the prosecutor
A trial in the Dutch criminal justice system is, in terms of its Criminal
Procedure Code,158 followed by the court’s judgment (vonnis), which
includes not only the conviction or acquittal but, in case of a conviction, the
sentence as well.159 However, before the judgment, the various parties get
the opportunity to address the court. The process is governed by s 311(1)
of the Code, which essentially provides as follows:
    ‘After the questioning of the suspect, and after the evidence by the witnesses and experts has
    been heard, the prosecutor may address the court; after reading his request, he hands it to the
    court. The request includes the sentence, if its imposition is requested. . . .’160
   A further statutory reference to the prosecutor’s ‘request’ (vordering) is
found in s 359(7) of the Code:
    ‘If a more severe punishment is imposed than that requested by the prosecutor, or if
    unconditional incarceration is imposed for a longer term than that requested, the judgment
    shall give the reasons leading to this decision in particular.’161
   The vordering is part of Dutch legal tradition. The prosecution performs
part of its role as representative of society by exercising this duty to request a
specific punishment.162 Despite not being bound by it,163 the judges cannot
simply ignore the request but have to explain any substantial departure in the
judgment.164 It should be mentioned here that, even though the sentence has
to be motivated in the judgment, in practice this motivation has become
almost meaningless.165
   The Prosecution Service (De Openbaar Ministerie) plays a particularly
important role in the Dutch criminal justice system. However, managing the
prosecution of crimes remains its primary function, and to that end it issues
directives on typical prosecution duties.Since requesting a specific sentence is
part of the prosecutor’s functions, the Openbaar Ministerie started issuing
guidelines to the prosecutors regarding the sentences that should be

  158    Wetboek van Strafvordering.
  159    Section 350.
  160    The original reads: ‘Nadat de ondervraging van de verdachte heeft plaatsgehad en de aanwezige
getuigen en deskundigen zijn gehoord,kan de officier van justitie het woord voeren;hij legt zijn vordering na
voorlezing aan de rechtbank over. De vordering omschrijft de straf en maatregel, indien oplegging daarvan
word geëist. . . .’
    161  The original reads: ‘Indien een zwaardere straf wordt opgelegd dan de officier van justitie heeft
gevorderd, dan wel een straf onvoorwaardelijk word opgelegd die vrijheidsbeneming van langere duur
medebrengt dan de officier van justitie heeft gevorderd,geeft het vonnis steeds in bijzonder de redenen op die
daartoe hebben geleid.’ The provision is intended to prevent surprize. The prosecutor’s request represents, in
principle, the upper limit of punishment that is to be imposed on behalf of society (cf P Osinga Transactie in
strafzaken: Een onderzoek naar de positie van de transactie in het strafrechtelijk systeem (1992) 36). It could have
affected public confidence in the judge’s role as ‘protector of the citizen’ if more severe punishment than the
prosecutor thought appropriate is imposed without explanation (cf Corstens op cit note 153 at 668).
    162  Osinga op cit note 161 at 36.
    163  H de Doelder in Van de Bunt et al op cit note 153 at 42–3.
    164  Osinga op cit note 161 at 36.
    165  Peter J P Tak The Dutch Criminal Justice System (1999) 74:‘[It] is generally known that this motivation,
required by art 359, (5) CCP, is pre-printed on the sentence form or flows easily from the word processor
when devising the verdict. . . .’ See also Cleiren & Nijboer op cit note 156 at 885 (especially due to a lack of
time); Corstens op cit note 153 at 666 (explanations such as ‘the imposed sentence is in accordance with the
seriousness of the offence and the circumstances under which it has been committed, and the person and
personal factors of the accused, as they appeared during the trial’ has long been accepted by the Hoge Raad,
despite objections from the practice and academia).
SENTENCING GUIDELINES FOR SOUTH AFRICA                                        877
requested.166 Initially, these guidelines were not particularly effective in
creating consistency, mainly because they allowed too much variation.167 By
1992, however, this had changed. Studies showed that judges were prepared,
in the interests of consistency, not to deviate much from the requests.168 In
this manner the guidelines of the Openbaar Ministerie gained a major
influence over judicial sentencing within a reasonably short time.169 It also
created more consistency in sentencing throughout the country. 170
   However, it did not stop here. By 1999, following an ambitious project
(called the Polaris project), 35 sentencing request guidelines had been
developed nationally, covering 125 offences (80 per cent of sentenced
cases).171 The guidelines operate in the following manner.172 The offender’s
basic offence is awarded a predetermined number of points (for example, 60
points in the case of breaking into a home). As the next step consideration is
given to reducing or increasing this number by a variety of mitigating or
aggravating factors affecting the basic offence. Each of these factors has also
been assigned a certain number of points. Thereafter a number of factors
affecting sentencing in general is taken into account. There would, for
example,be additional points if the victim is a public servant, or if the accused
operated in conjunction with others, or points subtracted if the victim
provoked the offender. If the offender has one previous conviction, the
number of points is increased by half, and doubled if there are more previous
convictions.173 The total number of points, called sanction points, determine
the sentence. Generally, one sanction point amounts to a fine of A22, or one
day’s imprisonment. However, imprisonment should only be requested if it
would amount to more than six months’ imprisonment (i e, if the sanction
points are more than 183). Otherwise a non-incarcerative sentence should be
requested, such as a fine or community service. The system also provides for a

   166   Van de Bunt et al op cit note 153 at 1–3.For the historical development,see S Faber ‘OM-richtlijnen:
geen nieuw verschijnsel’ in Van de Bunt et al op cit at 5–23. See also Osinga op cit note 161 at 79ff; Corstens
op cit note 153 at 217–8; J F Nijboer De taken van de strafrechter (2000) 45.
   167   Tak op cit note 165 at 77–8;Tak op cit note 156 at 175–6.The prosecutor was also allowed to deviate
from these directives without giving reasons.
   168   Osinga op cit note 161 at 37. J Knap ‘Rechtsgelijkheid bij de straftoemeting’ (1992) 71 Proces 161 at
162 estimated that sentences were about 25 per cent lower than the request.
   169   See also De Doelder op cit note 163 at 43.
   170   Tak op cit note 165 at 77. See also G Schurer & R van Loon ‘The Netherlands Adopts Numerical
Prosecution Guidelines’ in Tonry op cit note 155 at 170.
   171   Schurer & Van Loon op cit note 170 at 169. The guidelines were developed based on current
sentencing practices (can thus be seen as ‘descriptive’ — see note 30 above), with the use of a computer data
base: ‘The working group [developing the guidelines] contained two study groups, one to formulate new
guidelines and the other to test them. For more than a year, the first group collected all local and national
guidelines, which were then analysed, structured, and rewritten. The results were submitted to the testing
group, which checked the revised guidelines for content, internal consistency, and compliance with legal
standards’ — Schurer & Van Loon op cit 170–1.
   172   This discussion is mainly based on Schurer & Van Loon op cit note 170 at 169–74.See also Tak op cit
note 156 at 176–7; Tak op cit note 165 at 23. An interactive information system called BOS (Beslissings-
Ondersteuningssysteem) supports application of the guidelines in any given case — J de Hullu ‘De Plaats van
Databanken over de Straftoemeting in het Strafrechtelijk Sanctiestelsel’ (2002) Trema — Straftoemetings-
bulletin 1 11 at 16. Cf, for some of the latest and amended directives, M J A Duker and P M Schuyt
‘Straftoemeting’ (2002) 32 Delikt en Delinkwent 633 at 638–9. All the guidelines can to be found at [29 October 2003].
   173   Cf [29 October 2003].
878                                                     THE SOUTH AFRICAN LAW JOURNAL
conversion system when the number of points gets too high, especially in the
case of multiple offences.174 If a factor that the prosecutor considers important
in a specific case is not dealt with by the guidelines, it might be taken into
account if the prosecutor gives satisfactory reasons.
   It is too early for a proper assessment of the directives, but so far most
comments have been positive,stressing that the system is simple to operate.175

Actions taken by the judiciary
With all the activity by the prosecution, there has been considerable pressure
on the judiciary also to do something to improve sentence consistency.
Sentencing is, after all, a judicial function.176 Part of the pressure came from
the creation of a single economic market and efforts to harmonise the laws of
European countries.177 In the sphere of consistency in sentencing this led the
Council of Europe to draw up Recommendation R (92) 17 in 1996, which
referred specifically to sentencing orientation points and starting points.178
Developments in the Netherlands finally came to a head thus, Schoep
    ‘The National Conference of Presidents of the Criminal Chambers of the Courts of Appeal
    (NCPCCA)180 took up the responsibility and stated that they, the judiciary, should make
    efforts to promote consistency in sentencing. [In] 1998 a special commission on sentencing
    issues started formulating and issuing orientation-points for sentencing.’181
An important initial decision of the special commission (landelijke commissie
straftoemeting) was to divide the criminal cases into three groups:182
(1) The first group consists of those offences which the prosecutor may
       divert.183 The commission found that the judiciary had little need for

  174    Between 181 and 540 points,they are discounted by 50 per cent;above 540 points by 75 per cent.
  175    Not everybody is equally impressed. J M Reijntjes ‘Het openbaar ministerie: naar een nieuw
evenwicht’ (2002) 32 Delikt en Delinkwent 21 at 35 complains that some directives are ‘niksseggend,’ and
questions whether the necessary planning and control is in place, so that directives from which the courts
deviate substantially can be identified and scrapped.
   176   L Frijda ‘Eventuele Straatburgse aanbevelingen over straftoemeting’ (1992) 15 Trema 94 at 101.
   177   Nijboer op cit note 166 at 69.
   178   Frijda op cit note 176 at 94–7.Although one of the members of the committee of sentencing experts
is of the opinion that this recommendation has had little influence on sentencing in member states (cf
Andrew Ashworth ‘European sentencing traditions: Accepting divergence or aiming for convergence?’ in
Cyrus Tata & Neil Hutton (eds) Sentencing and Society: International Perspectives (2002) 219–36 at 223), this is
almost certainly not true of the Netherlands (cf J A W Lensing ‘Boekbesprekingen’ (2002) Trema —
Straftoemetingsbulletin 3 82 at 84 — a review of the aforementioned work; P A M Hendriks ‘Op weg naar
consistente straftoemeting: een reisverslag’ (2002) Trema — Straftoemetingsbulletin 2 57 at 58 — the
NCPCCA (see note 180 below) was guided in its desire for consistency by Recommendation R (92) 17);
R C P Haentjens ‘Over het motiveren van straffen’ (2002) Trema — Straftoemetingsbulletin 3 65 at 68.
   179   G K Schoep ‘Judicial co-operation in sentencing — shifting perspectives in the sentencing decision’
(2002) paper presented at the Second International Conference on Sentencing and Society, Glasgow, 27–29
June 2002.
   180   The landelijke overleg van de voorzitters van der strafsectoren van hoven en rechtbanken (LOVS).
   181   For more information, see Haentjens op cit note 178 at 65–71.
   182   This discussion is mainly based on Hendriks op cit note 178 at 58.
   183   Through what is called ‘transactie’. Transactie is a form of pre-trial settlement, in terms of which the
suspect undertakes to comply with the conditions as agreed upon with the prosecutor, in exchange for not
being prosecuted — Corstens op cit note 153 at 754. The most common condition is the payment of an
amount of money, very much like a fine. Other conditions are also possible, including payment of
compensation to the victim — Tak op cit note 156 at 156–7. Whether or not the suspect is actually guilty, is
immaterial. Because there is no trial, there is no conviction and no criminal record. For a detailed discussion
SENTENCING GUIDELINES FOR SOUTH AFRICA                                          879
        its own guidelines regarding this group of offences. The act (daad-
        component) is the main determinant of the sentence, and does not vary
        much from case to case. The influence of personal and personality
        circumstances is limited throughout. The sentence request by the
        prosecutor forms an important, and sufficient, orientation point.
(2) The second group of cases consists of those offences for which four
        years’ imprisonment or more are imposed. Since little information was
        available on these convictions,it was decided to set up a database of these
        sentences before further guidelines are prepared.
(3) The third group of cases consists of those offences falling inbetween the
        first two.This comprises a large section of the cases dealt with on a daily
        basis in the courts. It was decided to develop orientation points, or
        starting points (oriëntatiepunten), for these offences. Such oriëntatie-
        punten now exist for a substantial number of drug-related, violent,
        drunk-driving and property offences and it is the intention to issue
        more oriëntatiepunten.184 Much like the guidelines for the prosecutors
        the orientation points have been based on current sentencing
    The database for the second group of offences has been in operation since
1 May 2002.186 It has been made available to every member of the judiciary.187
Through the database judges are now able to determine how their colleagues
deal with comparable cases. In order to get a better perception of how judges
reach their decisions, for inclusion in the database they are required to give a
more complete judgment than has so far been done.188
    There has been some criticism of the database,some of it inspired by factors
of mostly local importance,189 but others by more basic considerations.190
Most commentators are, however, cautiously optimistic.191 The database is
still in its infancy, and only time will tell how well it will work.

of transactie see Osinga op cit note 161; see also Corstens op cit note 153 at 753. For a South African
perspective, see generally SA Law Commission Project 73: Sixth Interim Report on Simplification of Criminal
Procedure (Out of Court Settlements in Criminal Cases) (2002).
   184   Schoep op cit note 179. See also Hendriks op cit note 178 at 59.
   185   Hendriks op cit note 178 at 59; A F M Brenninkmeijer & G K Schoep ‘Gelijkheid van Straffen: Een
rechterlijk perspectief ’ (2002) Trema — Straftoemetingsbulletin 1 19 at 24
   186   Haentjens op cit note 178 at 65.
   187   Ibid. Access is via a system called ‘Porta Iuris’ — Editorial ‘Databank Straftoemeting’ (2002)
Trema — Straftoemetingsbulletin 1 1.An important practical aspect of the development of the database has been
the existence of a previously developed sentencing information system (see Haentjens op cit note 178 at
65–7).The database contains a special search function enabling the presiding officer to search for comparable
cases — M J A Duker and P M Schuyt ‘Straftoemeting’ (2002) 32 Delikt en Delinkwent 633 at 639.
   188   Haentjens op cit note 178 at 68–9. See, in connection with the current requirement for the
judgment, note 166 above.
   189   See,e g,G Knigge ‘Databank en strafmotivering:Catch-22?’(2002) Trema — Straftoemetingsbulletin 1
   190   Cf H K ter Brake ‘De Ideale Strafmotivering (overpeinzingen van een raadsman)’ (2002) Trema —
Straftoemetingsbulletin 3 79 at 80.
   191   De Hullu op cit note 172 at 18; Brenninkmeijer & Schoep op cit note 185 at 27 (the database is the
most important component of the development of a sentencing policy).
880                                                       THE SOUTH AFRICAN LAW JOURNAL
   Sentencing information databases, or systems, have been in operation in a
certain number of jurisdictions.It is,in conclusion,necessary to briefly look at
how they function in connection with sentencing consistency

    ‘A Sentencing Information System provides information about the range of penalties which
    have been passed by the court for a particular type of case in the past. The system allows the
    judge to enter information about the offence and the criminal history of the offender into
    the computer and the screen displays the range of penalties passed by the court for similar
   The advantage of a sentencing information system is that it provides
sentencers with accurate information about the sentencing practices of other
sentencers covered by the database.
   There has been a number of experiments with sentencing information
systems worldwide, but most of them ran out of steam. One system that has
been running for some time is administered by the Judicial Commission of
New South Wales, Australia.193 In addition to the Dutch database, such a
system has also recently been developed in Scotland.194 The staying power of
these latest initiatives still needs to be proved. Other systems have run into the
problems of cost, disinterested judges (who are often not computer literate)
and the fact that they tend to be driven by a few individuals, who lose interest
in the long run, or move on to other things.195
   Would a sentencing information system be of any substantial use in South
Africa? The advantages of knowing what one’s colleagues are doing and have
done are considerable. But the question has to be asked: Why did the Chief
Justice of New South Wales, the one jurisdiction where a sentencing
information system has been running for quite some time,196 find it necessary
to begin delivering guideline judgments? Clearly the system was not
sufficiently effective in producing consistency. This highlights the most
serious deficiency of sentencing information systems, namely that they
describe past sentencing practices, warts and all. There is no system to ensure
that only good sentencing judgments are included in the database.

The assessment in 2000 by the South African Law Commission regarding the
ills of our sentencing system remains valid. If anything, the current minimum

   192   Cyrus Tata et al A Sentencing Information System for the High Court of Judiciary of Scotland: Report of the
Study of the First Phase of Implementation and Enhancement (2002) 5.
   193   Cf New South Wales Law Reform Commission Report 79: Sentencing (1996) 14 (‘There is now
available to all sentencing courts, on a comprehensive basis, information about sentencing patterns through
the Judicial Commission’s Sentencing Information System . . . [T]he Commission considers that the greater
use of this information, which will inevitably come with the expansion of the database, will enhance
consistency’). See also Tata et al op cit note 192 at 7–8.
   194   Cf Andrew Ashworth ‘Sentencing’ in Mike Maguire et al (eds) The Oxford Handbook of Criminology
(2002) at 1106.
   195   Cf, in general, Tata et al op cit note 192.
   196   It was initiated in 1988 — Tata et al op cit note 192 at 8.
SENTENCING GUIDELINES FOR SOUTH AFRICA                                  881
sentence legislation has worsened the disparities and inconsistencies.197 In
addition to the arguments advanced by the Law Commission itself for its
finding that a Sentencing Council should be instituted in South Africa, it is
useful to consider the following conclusion that Tonry reached from the
American experience:
    ‘The strength of the sentencing commission approach is its combination of the sentencing
    commission, presumptive sentencing guidelines, and appellate sentence review.’198
    The importance of these three components is clear from research
conducted all over the world, and particularly in the countries highlighted
above. South Africa already has the ‘appellate sentence review’, in an
established appellate system on sentencing.To improve the sentencing system,
the other two components need serious consideration.
    The second component,namely sentencing guidelines,has been dealt with
at length in this article. In each of the jurisdictions considered disparity in
sentencing, with all its detrimental byproducts, has been the driving force
behind the move towards sentencing guidelines, regardless of the form that
these guidelines took. Without some point of departure within a rational
sentencing system, consistency is impossible to achieve. And, lest it be
forgotten in the sphere of sentencing, everyone ‘is equal before the law and
has the right to equal protection and benefit of the law.’199
    With respect to sentencing guidelines it is important, however, when
considering the Law Commission’s proposals,that one does not get caught up
in the mindset that only the American type of guidelines qualify as being
sentencing guidelines. In fact, there is much to say for the narrative approach
to guidelines of the English system, or even the points system of the Dutch
prosecution service, and it is submitted that they would produce a far better
‘fit’ in better with our legal tradition than the grid-like, more mechanical
American systems.
    The final component is the body providing the sentencing guidelines,
which may be called a sentencing commission or a sentencing council or
whatever else.Once again the different motivations for the institution of such
bodies have been canvassed at some length in this paper. The single most
important motivation is that courts are not in a position to establish a
complete system, since they deal with one case at a time. In considering this
matter, one also needs to consider the various functions of the proposed
Sentencing Council. Apart from the setting of guidelines it will also be
responsible for determining the value of fine units, making policy recom-
mendations on the development of community penalties (or other forms of
punishment), doing research (especially with respect to such matters as
sentencing trends and crimes figures), training judicial officers and producing

   197  Sections 51 and 52 of the Criminal Law Amendment Act 105 of 1997. For comment regarding the
increased risk of disparity, see S S Terblanche ‘Aspects of minimum sentence legislation: Judicial comment
and the courts’ jurisdiction’ (2001) 14 SACJ 1 at 19.
   198  Op cit note 36 at 283.
   199  Section 9(1) of the Constitution of the Republic of South Africa, Act 108 of 1996.
882                                                 THE SOUTH AFRICAN LAW JOURNAL
various publications.200 Currently most of these functions, although essential
for the proper operation of a sentencing system, are totally neglected in South
Africa. Without the establishment of some institution to effect these
functions, they are unlikely ever to be performed.
   The experience of the jurisdictions dealt with in this contribution proves
that, even within sophisticated criminal justice systems, eventually, proper
sentencing needs a sentencing guideline commission of some kind to guide
the sentence discretion.

        ‘The only person I ever met who saw the State was an elderly professor. ‘‘I have seen
    it,’’ he assured me solemnly. Did he really? Do we see the State, the Government, the
    Town, the Church, the Army, the University? Terms for these entities are repeated so
    often in political as well as mundane discourse that we begin to take them as referents for
    something tangible. But the whole is not more than the sum of its parts. Human
    interaction is no more and no less than people interacting with one another,often in very
    small groups, many of them composed of only two people and lasting a very short time.
    Those other grand terms are abstract symbols. Human beings do not interact with
    symbols, nor are they injured by them. They interact with — are helped, reared, robbed,
    raped, killed by — other human beings, who relate to them in various patterns of
    equality,subordination,and superordination.Persistently overlooking ineluctable human
    agency and blaming the symbol becomes a technique of amnesty for individual
    perpetrators and a continuing invitation to impunity.’
        W Michael Reisman Law in Brief Encounters (1999) 4.

   200  South African Law Commission Report: Sentencing (A New Sentencing Framework) (Project 82) (2000)
para 3.2.4ff.