A Pyrrhic victory?

Mandatory and minimum sentences in South Africa

Julia Sloth-Nielsen & Louise Ehlers

Occasional Paper 111, July 2005

Introduction

The period preceding Parliament’s extension of South Africa ’s mandatory sentencing laws in April 2005 saw intense lobbying, advocacy, and public debate about the efficacy and desirability of minimum sentences for serious crimes. Although the extension of the period of operation of the Criminal Law Amendment Act 105 of 1997 – which introduced prescribed sentences for specified serious offences into our legislative framework – had occurred on two previous occasions, this happened without much fanfare or publicity. The sources expressing more fundamental concern were the non-governmental sector, the Office of the Inspecting Judge on Prisons, and, unsurprisingly, the judiciary.1 This paper examines some of the arguments that have been raised for and against the present sentencing regime, and highlights the need for comprehensive sentencing reform in South Africa .

The paper also examines some key questions: Are mandatory minimum sentences constitutional? Have they deterred or prevented crime? Do they afford better protection to victims? What is the relationship between minimum sentences and prison overcrowding? Finally, the paper questions whether South Africa needs a more comprehensive sentencing reform strategy.

Background to mandatory sentences in South Africa

Sentencing in South Africa has traditionally been the preserve of the judiciary. With regard to statutory offences, like drunken driving or possession of unlicensed firearms, sentencing officers are bound by the form of punishment prescribed in the applicable act, as well as by the maximum punishment set out there. In respect of common law crimes, such as theft, robbery and rape, the possible sentences are not prescribed by any legislation. The exercise of judicial discretion is circumscribed only by the maximum jurisdiction of the applicable court. The punishment jurisdiction of district and regional courts is determined by section 92 of the Magistrates’ Courts Act 32 of 1944, and presently this is a maximum of three years imprisonment in the case of the former, and 15 years in the case of the latter. High Court sentencing jurisdiction is not limited and may extend to the maximum sentence now permissible in South Africa’s legal system, namely life imprisonment.

Judges and other sentencing officers have traditionally resisted interference with their sentencing discretion, which they regarded as a fundamental aspect of judicial independence. In 1971, the Viljoen Commission of Inquiry into the Penal System expressed opposition to legislative interference with judicial sentencing discretion in the form of prescribed sentences, recommending the removal of certain sentences (for corrective training and the prevention of crime) from the statute book.

A source of fierce public debate during the early years following South Africa’s transition to democracy was the perceived leniency in punishing serious offenders, coupled with the perception that offenders were not serving substantial enough portions of their sentences due to a lax parole policy. The public was also concerned about the nature and severity of sentences for heinous crimes after the abolition of the death penalty.2

Against this backdrop, the Minister of Justice and Constitutional Development in 1996 appointed a project committee of the South African Law Reform Commission,3 with the brief to investigate all aspects of sentencing, including the desirability of legislation on minimum and maximum sentences.

The committee operated from June 1996 to March 1998 under the leadership of Judge Leonora Van der Heever, and during 1997 produced an Issue Paper No. 11 entitled (Project 82) Sentencing: Mandatory minimum sentences.

Issue Paper 11 (Project 82), Sentencing: Mandatory minimum sentences

The project committee set about identifying the critical issues in the sentencing arena. As a point of departure, it took the view of Ashworth that:

sentencing is the stage after the determination of criminal liability and may be characterised as a public, judicial assessment of the degree to which the offender may rightly be ordered to suffer legal punishment. It is therefore important not to lose sight of the fact that the values which society wishes to uphold should be those which inform any reform in sentencing.4

Issues identified by the committee included the following:

Lack of uniformity and consistency in sentencing

Illustrating the extent of the lack of consistency in sentencing, the Issue Paper cites the facts in S v Young,5 where two learned judges gave careful consideration to the same issues arising out of a set of agreed facts, but arrived at diametrically opposed conclusions.

Nairn argued at the time that the nature of the sentencing procedure made this type of outcome virtually inevitable, given the fact that, although the course of the trial is determined by clearly defined rules of law, the approach to sentencing is largely left to chance.6 Nairn was of the view that in the absence of clearly articulated guidelines, uniformity in sentencing remained unattainable.

In the light of the above, it was submitted that:

it is no longer enough to list aggravating and mitigating factors and then to move straight on to a generalised conclusion. The lower courts are in desperate need of a comprehensive set of principles, which can be used as basic guides. Ideally these principles should be formulated by the Appellate Division so as to ensure their uniform application throughout the country. Upon these principles could be built a comprehensive body of sentencing law – our only road out of the quagmire.7

Main characteristics of punishment in SA

• The project committee was of the view that mandatory sentences could not be discussed without reviewing sentencing practices in South Africa as a whole.
  
  

• Reviewing the justification for imposing punishment, the Issue Paper, after defining punishment as the sanction of the criminal law, reported that there appeared to be general consensus on the two outstanding characteristics of punishment, namely: the intentional infliction of suffering upon an offender on account of the commission of a crime; and the expression of the communities’ condemnation and disapproval of the offender and his/her conduct.8
  
  

• In articulating the aims of punishment recognised by the courts in South Africa, the Issue Paper quoted S v Khumalo9 where it is stated that: “In the assessment of an appropriate sentence, regard must be had inter alia to the main purposes of punishment mentioned by Davis AJA in R v Swanepoel 1945 AD 444 and 445, namely deterrent, preventative, reformative and retributive … .” A popular view is that the ultimate aim of punishment is to protect the community against crime. According to the Issue Paper, the differences in opinion arise as to the best method of achieving this. One method is by directly incapacitating the offender, for example, by imposing the death penalty. Alternatively, protection might be achieved by indirect prevention, where the aim is to persuade the offender to cease his/her activities voluntarily.

The Issue Paper provided a synopsis of the existing legislative framework for sentencing in South Africa at the time of drafting, detailing the specific sentences provided for in the Criminal Procedure Act 51 of 1977. As far as judicial discretion in sentencing was concerned, the Issue Paper suggested that the legislature’s primary task – in addition to defining what conduct would be criminal and providing a threat of punishment – was also to prescribe the nature of the punishment that may be imposed and the maximum punishment that may not be exceeded. The Issue Paper stated:

[I]t is generally accepted that the South African Courts have a discretion to determine the nature and extent of the punishment to be imposed within this framework.10

Criticism of the control over sentencing exercised by appeal and review courts is that a court of appeal will only reverse a decision of a trial court if it appears that the trial court has exercised its discretion in an improper or unreasonable manner.11 While a number of tests have been developed to determine when it is appropriate to interfere with a trial court’s decision, it was argued that the tests are vague and imprecise.12

The introduction of mandatory sentences

Owing to previous concerns that the legislative framework left too wide a discretion to the courts, a number of attempts had been made to limit sentencing discretion by providing for mandatory minimum sentences. By way of illustration, the Issue Paper cited the mandatory imposition of corporal punishment under certain circumstances, introduced in 1952. Also, in 1959 compulsory imprisonment for the prevention of crime and imprisonment for corrective training were introduced in certain circumstances. Mention was made of the Abuse of Dependence Producing Substances and Rehabilitation Centres Act 41 of 1971, which also contained a number of mandatory sentences.

Strong opposition was noted to previous provisions containing mandatory minimum sentences, notably in S v Toms; S v Bruce where Chief Justice Corbett observed that:

… the imposition of a mandatory minimum prison sentence has always been regarded as an undesirable intrusion by the Legislature upon the jurisdiction of the courts to determine the punishment to be meted out to the person convicted of a statutory offence and as a kind of enactment that is calculated in certain instances to produce grave injustice.13

It was also argued, however, that South African courts have often been criticised for adopting an intuitive approach to sentencing, with sentences being passed on an unscientific basis.

At the time that the Issue Paper was written, the point of departure in the sentencing process was the well known dictum in S v Zinn:14 that is, that the triad consisting of the crime, the offender and the interests of society be considered. This triad has been criticised as vague, elementary and unsophisticated; further, the role of the victim is not emphasised.

The position was taken that failure by the legislature to provide a clear and unambiguous legislative framework for the exercise of the sentencing discretion, coupled with failure by the courts to develop firm rules for the exercise of the sentencing discretion; and further failure by the courts and the legislature to give firm guidance as to which sentencing theories or aims carry the most weight, brought much uncertainty and inconsistency into the sentencing process in South Africa.15

The international perspective

The Issue Paper noted that a number of factors had in the years prior to the compilation of the Issue Paper sharpened the debate internationally about sentencing policy. These factors included rising crime rates, prison overcrowding, fiscal crises, loss of faith in the offender treatment paradigm, concern that just deserts be delivered, and the need for public protection against dangerous offenders.

As in South Africa, the rise in violent crime and growing public demand for harsher and more certain punishment resulted in many countries, including England, the United States (US) and Sweden, instituting sentencing reforms. The Issue Paper saw these reforms as primarily a response to criticism of ineffective rehabilitation attempts. It noted, however, that a range of differing goals had influenced various sentencing reforms, including reducing the disparity that results from discretionary sentencing, increasing sentencing fairness, establishing truth in sentencing, and balancing sentencing policy with limited correctional resources.

Consideration was given to sentencing reforms in a range of countries. The US was particularly influential, where sentencing guidelines developed by an independent sentencing commission represented the dominant approach to sentencing reform. Many states have replaced indeterminate sentencing with structured sentencing schemes, such as determinate sentencing,16 presumptive sentences, mandatory minimum penalties, and sentencing guidelines.17 The Minnesota Guidelines (discussed further at p 6) are probably the most well known; by 1997 there were over 60 criminal statutes containing mandatory minimum sentences in the US Federal code.

A number of justifications were advanced for the enactment of this type of legislation. These included:

• retribution/just desserts – it is argued that the punishment should fit the severity of the crime, and that past leniency should be corrected;
  
  

• incapacitation – it is vital to ensure the incapacitation of serious offenders to protect the community;
  
  

• disparity – mandatory minimum sentences are held to reduce unwarranted disparity in sentencing; and
  
  

• inducement of co-operation – mandatory minimums may help induce defendants to co-operate with authorities.18

Developments in Sweden, Germany, England and Wales, Canada, and Greece were also explored.

Possible options for sentencing reform

In conclusion,19 the Issue Paper provided a range of possible options for sentencing reform:

• Presumptive sentencing guidelines: To set up a sentencing commission to develop sentencing guidelines in respect of certain offences, which guidelines would give specific principles (for example, the severity of the offence and the accused’s prior criminal record) to determine the presumptive sentence. The court is allowed to depart from the presumptive correct sentence only if special circumstances exist.
  
  

• Voluntary sentencing guidelines: These require the development of sentencing guidelines; but these guidelines do not by law have to be followed – they simply guide the courts in the exercise of their discretion.20
  
  

• The adoption of legislative guidelines that assist in determining the choice and length of the punishment: Based on the Swedish model, this option entails that the legislature determines the nature of punishment (e.g. community-based or prison) and the penal value attributed to the particular offence.
  
  

• The enactment of principles of sentencing, including guidelines that determine the imposition of imprisonment: Based on the proposals of the Canadian Sentencing Commission, provision is made for principles governing the determination of the sentence: that is, that the sentence should be proportionate to the gravity of the offence and the degree of responsibility of the offender for the offence. In addition, a number of factors are listed which the court has to consider in determining the sentence. These include, but are not limited to: aggravating and mitigating circumstances; the need for consistency in sentencing; the need not to impose excessive sentences; the fact that imprisonment should not be imposed solely for the purposes of rehabilitation; the circumstances under which imprisonment should be imposed; and consideration of the aims of punishment.21
  
  

• The enactment of presumptive sentencing guidelines to guide the imposition of custodial and non-custodial sentences: Presumptive sentencing guidelines are statutory orders that impose a predetermined sentence range to the judge. Although presumptive guidelines are statutory in nature, they do allow the continued existence of sentencing discretion for the judge under certain circumstances.
  
  

• The enactment of mandatory minimum sentences combined with the discretion to depart from the sentences under certain circumstances: This option implies the enactment of a mandatory minimum sentence (e.g. 15, 20 and 25 years for a first, second, and third conviction respectively), coupled with discretion to depart from the prescribed sentence if special circumstances exist. In these cases, the court is required to record the circumstances of the case and to give written reasons for the departure from the prescribed sentence. (The most well known example, the Minnesota Guidelines, is discussed further at p 6.)

The Issue Paper recommended that the issues raised be discussed and debated thoroughly before any particular direction was taken. Based on the outcomes of these discussions, legislation in respect of sentencing would be proposed. Unfortunately, however, by the closing date for public comment, the Criminal Law Amendment Act 105 of 1997 – which introduced mandatory minimum sentences in South Africa – was almost already finalised in Parliament.22

The work of the project committee in relation to mandatory minimum sentences was therefore superseded by events, and the committee chaired by Judge of Appeal Van der Heever completed its term of office without consolidating its work in a discussion paper or in draft legislative proposals.23

There is no doubt that the Law Commission process was, in effect, bypassed in the rush to enact legislation giving effect to stricter sentencing laws. The minimum sentencing legislation must, however, be seen in the context of a range of other legislative and policy shifts that occurred around this time to deal with the perception that government was not taking the crime problem seriously. These included harsh new bail laws, legislation dealing with organised crime and the criminalisation of gangs, and eventually (subsequent to the enactment of minimum sentencing laws), the downgrading of the National Crime Prevention Strategy (NCPS) in favour of a far tougher, no-nonsense approach to dealing with crime.

There was little public consultation about the proposed legislation or its form.24 Members of another South African Law Reform Commission’s project committee that was drafting legislation for a new juvenile justice system, learnt by chance of the proposal to include children within the reach of minimum sentencing laws.25 The rush was such that the legislation was badly formulated, as has frequently been noted by courts and academics.26 It was, after all, intended to be only temporarily in operation; a short-term emergency political measure to allay public fear.

However, as suggested in the concluding part of this paper, if a wider public debate about sentencing reform is to take place now, the initial study on mandatory sentencing undertaken by the South African Law Reform Commission may again be relevant, particularly as regards the various possible law reform options cited in the concluding chapter of the Issue Paper.

A new project committee to investigate sentencing reform was nevertheless appointed by the Minister of Justice and Constitutional Development in 1998. The work of this committee culminated in the 2000 publication of a report entitled Sentencing: A new sentencing framework. The Report was accompanied by a proposed Sentencing Framework Bill, to be discussed later in this paper at page 15.

Structure of the minimum sentences legislation contained in Act 105 of 1997

Offences to which the sentences apply

The Criminal Law Amendment Act 105 of 1997 applies to offences committed after the date on which it came into operation, namely 1 May 1998. It lists some of the most serious offences – committed in circumstances thought to render the perpetrator especially blameworthy, and attracting particular public concern at the time – as those which would qualify for the prescribed minimum sentence. These include:

• Murder, where:

  • it is planned or premeditated;

  • the victim was a law enforcement officer performing his or her functions as such;

  • a person has given or was likely to give material evidence in certain specified cases;

  • the death of the victim was caused in the course of rape or robbery with aggravating circumstances; or

  • the offence was committed by a person or group of persons or syndicate acting in furtherance of a common purpose or conspiracy.27

• Rape, where it is committed:

  • in circumstances where the victim was raped more than once, whether by the accused of by any co-perpetrator or accomplice;

  • by more than one person, where such person acted in the execution or furtherance of a common purpose or conspiracy;

  • by a person who has been convicted of two or more offences of rape but has not yet been sentenced in respect of such convictions; or

  • by a person, knowing that he has HIV/AIDS.28

• Second, rape where it is perpetrated upon:

  • a girl under the age of 16;

  • a physically disabled woman who, due to her disability, is rendered particularly vulnerable; or

  • a mentally ill woman.29

• The third instance is when a person is convicted of a sexual assault inflicting grievous bodily harm on the victim.30

Other offences also targeted for prescribed minimum sentences include, among others: specific offences involving corruption, extortion, fraud, forgery and uttering, or theft in certain circumstances; dealing in drugs in certain circumstances; dealing in arms or ammunition in certain circumstances; robbery involving the taking of a motor vehicle; robbery with aggravated circumstances; and certain offences committed by law enforcement officials.31 The sentencing scheme (or grid) envisaged by the Act is reflected in a summarised form in Box 1.

Departure from the prescribed sentence in ‘substantial and compelling circumstances’

Section 51(3)(a) of the Act provides as follows:

If any court referred to in ss (1) or (2) is satisfied that substantial and compelling circumstances exist which justify the imposition of a lesser sentence than the sentence prescribed in those subsections, it shall enter those circumstances on the record of the proceedings and may thereupon impose such lesser sentence.

The term ‘substantial and compelling circumstances’ was not previously found in South African law. It appears to have been borrowed from American sentencing practice, as found for instance in the Minnesota context.33 In Minnesota, however, if substantial and compelling circumstances are found to exist, a departure from the guideline sentence can result in a harsher or a more lenient sentence being imposed.34

The Minnesota Guidelines contain a list of factors that may not be considered in the decision as to whether departure is justified, including race, sex, educational background, and marital status. Equally, a list of factors is provided indicating when substantial and compelling circumstances might be found, although this is not a closed list of reasons. The list includes the fact that: the victim was the aggressor in the incident; the offender played a minor role in the commission of the offence, or participated as a result of coercion or duress; the offender as a result of physical or mental impairment, lacked substantial capacity when the offence was committed (but the voluntary use of drugs or alcohol may not be considered under this heading); and, finally, any other substantial ground which tends to excuse the offender’s culpability.

Van Zyl Smit expresses the view that a noteworthy feature of the circumstances that may, and may not, be considered is that they all focus on features which relate to the gravity of the offence itself, and the blameworthiness of the offender in respect of that offence.35 Further, especially insofar as departures downwards are concerned, this means that sentences will not be grossly disproportional to the specific offence (the principle of ‘limiting retributivism’).

The legislature’s adoption of the ‘substantial and compelling circumstances’ standard was undoubtedly deliberate: a clear effort to ensure the ultimate constitutional validity of Act 105 of 1997; inflexible mandatory sentences risk producing disproportionate sentencing outcomes in hard cases. Canadian and Namibian jurisprudence in existence at the time that South Africa’s legislation was being drafted had concluded that a mandatory minimum sentencing provision that might reasonably result in a sentence which is grossly disproportionate to the crime, would offend constitutional principles.36

In one of the first cases in which the meaning of the phrase ‘substantial and compelling circumstances’ was considered, the presiding judge said that:

for ‘substantial and compelling’ circumstances to be found, the facts of the particular case must present some circumstance that is so exceptional in nature, and that so obviously exposes the injustice of the statutorily prescribed sentence in the particular case, that it can rightly be described as ‘compelling’ the conclusion that the imposition of a lesser sentence than that prescribed by Parliament is justified.37

Subsequent cases, however, did not follow this approach. In S v Malgas, the Supreme Court of Appeal disagreed that exceptionality was the true criterion for departure from the minimum sentence.38 The Court stated that the legislature had deliberately refrained from giving guidance as to which circumstances would or would not be regarded as substantial and compelling, and therefore sufficient to justify departure from the prescribed sentence.39 The Court was of the view that it would be an impossible task to attempt to catalogue the circumstances, or combinations of circumstances, which could rank as substantial and compelling, or those which could not.40

However, it had to be borne in mind that the legislature had specifically prescribed the sentence that would be ordinarily appropriate, so it would be inappropriate to justify departures from the prescribed sentence based on “spurious rationalisations or the drawing of distinctions so subtle that they can hardly be seen to exist”.41

The Court’s view was that the legislation had limited, but not eliminated, the Court’s discretion in imposing sentences for the crimes covered by the Act. In the absence of weighty justification (also put as “truly convincing reasons”), the Court held that these crimes should elicit a “severe, standardised and consistent response from the courts”.42

Nevertheless, most would argue that different approaches to what amounts to substantial and compelling circumstances still proliferate, which in turn entrenches sentencing disparity. This argument is dealt with more fully below.

Constitutional challenges

It was inevitable that the minimum sentencing legislation would ultimately be subject to constitutional scrutiny. The leading case in which this review occurred was S v Dodo.43 The legislation was challenged, among other reasons, on the basis that interference with the Court’s sentencing powers by the legislature breached the separation of powers doctrine, as a criminal trial before a court requires an independent judiciary to weigh and balance all factors relevant to the crime, the accused, and the interests of society in the sentencing process.44

The Constitutional Court linked the separation of powers argument to the accused person’s right not to be punished in a cruel, inhuman, or degrading way.45 The Court was of the view that little can be gained by examining South African jurisprudence on mandatory sentencing prior to 1994. In the new constitutional era, parliamentary sovereignty can be subjected to constitutional scrutiny.

The Court pointed out that both the legislature and the executive share a justifiable interest in sentencing policy (including matters such as the severity of sentences and the protection of the law-abiding public), as well as the execution of sentences (for example, the availability and cost of prisons). The Court said that the executive has a general obligation to ensure that law-abiding persons are protected (if needs be through the criminal law) from persons who are bent on breaking the law, and that this obligation is even more acute in regard to crimes involving violence against bodily integrity. Furthermore, it increases with the severity of the crime.50 However, the legislature’s powers are not unlimited and cannot wholly exclude the power of courts to apply and adapt a general principle to the individual case.51

The Court affirmed that the relevant inquiry is whether the statute concerned may result in a grossly disproportionate sentence.

The concept of proportionality goes to the heart of the inquiry as to whether punishment is cruel, inhuman or degrading, particularly where, as here, it is almost exclusively the length of time for which an offender is sentenced that is in issue.52

Where a lengthy sentence imposed to serve the aims of generally deterring others from committing offences bears no relation to the gravity of the actual offence committed in a specific instance, “the offender is being used essentially as a means to another end and the offender’s dignity is assailed”.53 However, the provision enabling departure from the prescribed sentence where substantial and compelling circumstances are found to have existed, makes it plain that disproportionality can be avoided and that courts are not forced to act inconsistently with the constitution. In addition, the Constitutional Court could not find any violation of the accused’s fair trial rights in section 35(3)(c) – the minimum sentences legislation in no way alters the character of the High Court, or detracts from it being an ‘ordinary court’ as provided for in that section.

Has minimum sentencing had the desired impact on prison conditions?

Several commentators, notably the Inspecting Judge of Prisons, have ascribed worsening prison overcrowding to the impact of minimum sentences. Writing in The Advocate in April 2005, Judge Fagan says:

The effect of the minimum sentence legislation has been to greatly increase the number of prisoners serving long and life sentences. It has resulted in a major shift in the length of prison terms as indicated in … [Figure 1].54

Judge Fagan records that sentences of seven years and less showed little change from 1997 (67,535) to 2004 (67,483), while sentences of more than seven years increased rapidly from 1997 (29,376) to 2004 (67,081). Life sentences increased from 638 in 1997 to 5,511 on 30 September 2004. He notes that in April 1998 – immediately before the implementation of the minimum sentence legislation – only 18,644 (19%) of the sentenced prisoners were serving a term of longer than ten years. This has since increased to 49,094 (36%) (see figures 2 and 3).

Judge Fagan asserts further that the numbers continue to rise. Writing in December 2004, he provides the latest available figures (at 30 September 2004) showing that there were then 5,511 prisoners serving life sentences compared to 4,460 12 months earlier; and 43,583 prisoners were serving sentences longer than ten years compared to 40,056 in September 2003. As indicated in Figure 4, the sentenced prisoner population has increased by 28,801 prisoners since April 2000, despite about 7,000 being released on parole in September 2003. Judge Fagan argues that with a growth rate of more than 7,000 prisoners a year, such inhumane conditions will result that mass releases will be required periodically (see Figure 4).

Lukas Muntingh of the Civil Society Prison Reform Initiative illustrated the rise in sentenced prisoner profiles in Figure 5.55

Judge Fagan’s views are supported by the Department of Correctional Services, which is also concerned about the burgeoning prison population. The Democratic Alliance Spokesperson on Justice has also publicly linked the minimum sentencing laws to overcrowding of prisons.56

Writing on prison overcrowding, Steinberg makes the point that it is “somewhat baffling that parliament passed the minimum sentencing provisions apparently without thought to the effect on prison volumes …”.57 He maintains that there is “abundant international evidence that a sudden and sustained increase in sentences for serious crimes will inevitably lead to an increase in prison numbers”.58 Steinberg cites, for example, the US where mandatory minimum sentences were introduced in the late 1970s and early 1980s. From 1980 to 1995, the US prison population grew by 242%. The generally accepted reason for this growth is the lengthening of prison sentences, the decreased possibilities of parole, and policies mandating incarceration for growing numbers of offences.59

It is an unassailable reality that the sentenced prison population in South Africa has increased rapidly since 1998. Moreover, the evidence is overwhelming that a significantly larger proportion are serving long terms of imprisonment – with the number of prisoners serving sentences of more than ten years having quadrupled from 10,000 to 40,000 in the past nine years.60

Nevertheless, critics argue that it cannot be conclusively shown that the increase in long-term and life sentences is necessarily due to the implementation of the minimum sentences legislation. It could, they assert, simply be due to a general increase in the prevalence of serious crime, or to a generally more punitive and intolerant mood among judicial officers. It could even be the result of better police clearance rates for serious offences.

The impact of increased jurisdiction of the lower courts at around the same time may also have to be factored in. For instance, the maximum sentence that a regional court could impose was increased from ten to 15 years, at the same time that prescribed sentencing laws were introduced. Yet another possible cause is the impact of the 2000 South African Police Services’ (SAPS) National Crime Combating Strategy (NCCS), which focused on selected crime areas with highest recorded crime levels, and directed police resources at these areas in the form of high-density search and seizure operations. Many more people were arrested, in 2001; the National Prosecuting Authority’s case intake went from around 500,000 to 1,1 million. Many of these cases resulted in arrests and awaiting trial prisoners.

Also, the split procedure entrenched in section 52(1) of Act 105 of 1997 created bottlenecks for awaiting trial prisoners as cases were delayed while records were typed and dates sought. The procedure requires regional courts that hear cases involving an offence in Part I of Schedule 2, and who convict an accused but prior to sentence are of the opinion that the punishment may exceed the jurisdiction of the regional court, to refer the matter to the High Court for sentence. It is not uncommon for a judge to send the matter back to the regional court magistrate for clarification, thus only adding to the delay.

A case-by-case analysis would be needed to establish conclusively the link between the implementation of the legislation and the otherwise rather persuasive statistics shown above – such analysis was done during the South African Law Commission’s investigation into sentencing.61 In its preliminary investigation (conducted shortly after the coming into operation of the Act) it was found that sentence severity for rape and robbery with aggravating circumstances committed after implementation of the Act significantly increased sentencing severity. Additionally, it was found that there was a large increase in the percentage of life sentences given for rape and murder after the implementation of the Act, but that the increase for murder was not significant.62 Any future analysis would have to investigate first whether the sentences were imposed pursuant to Act 105 of 1997, and second, whether in the absence of the prescribed sentence it is likely that a shorter or less severe term of imprisonment would have been the likely sentence. Since the provenance of the sentence does not appear on the detention warrant, this would require tracking actual cases – an intensive and potentially vast endeavour.

Furthermore, proper docket analysis is required before it can be proved or disproved whether any improvements in police investigative capability account for the increased numbers of prisoners serving long-term or life sentences.

Has the legislation prevented or curbed crime?

It is routinely noted that the impact of harsher sentencing regimes on general deterrence of crime is difficult to isolate and measure. Writing from an international perspective, Tonry noted that:

The evidence is clear and weighty, that enactment of mandatory penalty laws has either no deterrent effect or modest deterrent effect that soon wastes away. Equally clear and consistent are findings that mandatory minimum laws provoke judicial and prosecutorial stratagems, usually by accepting guilty pleas to other non-mandatory penalty offences or by diverting offenders from prosecution altogether that avoid their application.63

However, despite the repeated extension of an Act that was originally intended to be in place for only two years, it is difficult to find substantive evidence in a South African context which shows that the introduction of a new penal regime in 1998 has had any general deterrent effect – or even that it has had the effect of reducing crime.64 Instead, statistics suggest an uneven change in reported crimes.65

According to data analysed by Scharf and Berg66 from 1997/98, levels of recorded crime rose steadily, beginning to level off in 2000/01, dipping in 2001/02, but then rising again in 2002/03.67 It must of course be noted that a number of factors mediate the interpretation of data such as this. These factors include levels of reported crime that actually occur, means of data collection and recording, and shifting categories and methods of classification of offences. Nevertheless, the data appears to indicate a gradual decrease in violent crime as a ratio per 100,000 of the population from 1994/95 to 1997/98, after which the rates rose sharply.

For the purposes of this analysis, violent crime includes murder, attempted murder, rape, attempted rape, assault with intent to do grievous bodily harm, common assault, aggravated robbery, other robbery, and malicious damage to property. Some, albeit a limited number, of these offences are targeted by the mandatory sentencing laws.

By way of illustration, Scharf and Berg68 provide a table reflecting the incidence of crime per 100,000 of the population (see Table 1).

A more nuanced analysis per crime category revealed that in some respects, violent crime had decreased – murder being a case in point. However, aggravated robbery statistics rose, as did cash-in-transit robberies;69 both offences for which mandatory sentences might well be imposed.

Altbeker comes to similar conclusions with reference to the latest statistics released by the SAPS, noting, however, that methodological problems are inherent in collating and interpreting crime statistics, and that there are widely recognised discrepancies in reporting the rates of different crime categories. Murder is generally accepted as being a highly reported crime (there is after all a physical body to account for), whereas robbery is viewed as being largely under-reported. Altbeker notes that between 1996/97 and 2003/04, the incidence of murder per 100,000 of the population decreased steadily from 62.8 to 42.7.70 Indeed, a continued decline in the incidence of murder from 1994 onwards illustrates that the abolition of the death penalty in 1995 did not contribute to an increase in this offence, contrary to what many may believe.

Similarly, since a downward trend was already in evidence by the time that the minimum sentencing law was introduced, and there was no significant increase in the decline rate subsequent to the introduction of the minimum sentencing provisions, the decrease after this cannot be attributed to any intervening legislative changes.

During the same time period, however, the incidence of aggravated robbery increased from 218.5 of the population to 288.5.71 But then hijacking – which has dropped substantially since 1998 – is also a very highly reported crime (mainly due to insurance reasons), and this is an offence that is specifically included in Part II of Schedule 2 of the minimum sentences laws (the relevant clause mentions robbery involving the taking of a motor vehicle).72 It can therefore be concluded that, at present, there is little in the way of reliable evidence that the new sentencing law has reduced crime in general, or that the commission of the specific offences targeted by this law has been curbed.

A further argument in connection with the function and outcome of mandatory sentences was provided by Altbeker at the roundtable on minimum sentences hosted by the Open Society Foundation in January 2005.73 Altbeker avers that there are three transmission mechanisms through which custodial sentences might be argued to reduce crime, namely: rehabilitation; incapacitation; and deterrence. He notes that there is little evidence that the time prisoners spend in jail is effective in rehabilitating them; rather, there is more evidence to the contrary. Indeed, there is anecdotal evidence from members of the Department of Correctional Services that prisoners serving extremely long terms of imprisonment – such as those prescribed as mandatory sentences – have nothing to hope for. The prospect of release is so far off that they are consequently less amenable to any rehabilitative efforts.74

This leads to the second point made by Altbeker, namely: that the use of prison to incapacitate offenders breaks down when prison sentences are too long, as prisoners are kept behind bars well past the age at which most criminologists expect them to continue offending. After this point, they are consuming scarce prison space without doing much to reduce crime, since those behind bars would in any case be less likely to offend.75

Altbeker concluded by pointing out that evidence suggests that the preventative effect of a 1% increase in the certainty that an offender will go to prison is far more effective than a 1% increase in the length of a sentence.76 Thus, from a crime control perspective it is more efficient to use prison space for more people sentenced to shorter periods, than for fewer people sentenced to longer terms.77

Has the legislation promoted consistency in sentencing?

It must be borne in mind that the elimination of inconsistent and apparently widely diverging sentencing practices was a key objective underlying the introduction of the prescribed sentencing regime. As Prof. S S Terblanche has pointed out:

[t]he lack of consistency in sentencing is a major problem in South Africa, as it is in other countries where sentencers have largely unfettered discretion in imposing sentence.78

Terblanche asserts, however, that the minimum sentences legislation has, if anything, worsened the disparities and inconsistencies that prevail in relation to the offences targeted by the law.79 Certainly, despite assertions from judges and magistrates that the prescribed sentences require them to treat all those guilty of a particular type of crime in the same way, irrespective of differing circumstances, newspaper reports abound of apparently severe cases in which judges have found room to depart from prescribed minimum sentences.80

In a submission to the Minister of Justice and Constitutional Development in response to the calls for input prior to the extension of Act 105 of 1997, the Western Cape Consortium on Violence Against Women argued further that:81

… the Supreme Court of Appeal cases of S v Mahomotsa,82 and S v Abrahams83 added a gloss on the ‘substantial and compelling’ test that has resulted in confusion in the jurisprudence, and perhaps a change in the Malgas test …

In S v Mahomotsa, in discussing the interpretation of ‘substantial and compelling circumstances,’ Mpati JA stated:

“Even in cases falling within the categories delineated in the Act there are bound to be differences in the degree of their seriousness. There should be no misunderstanding about this: they will all be serious but some will be more serious than others and, subject to the caveat that follows, it is only right that the differences in seriousness should receive recognition when it comes to the meting out of punishment. As this Court observed in S v Abrahams, ‘some rapes are worse than others and the life sentence ordained by the Legislature should be reserved for cases devoid of substantial factors compelling the conclusion that such a sentence is inappropriate and unjust’.”84

The wording in the latter half of this paragraph seems to change the presumption such that rather than justifying departures from mandatory minimum sentences, judges should justify failures to depart therefrom.

Indeed, this was the interpretation that was seized upon by the court in S v G85 where the court found that although the mitigating factors could not be said to amount to substantial and compelling circumstances, the cases of Abrahams and Mahomotsa required the court to find that the rape in question is one of the most serious manifestations of rape before imposing a life sentence on the accused.

These arguments convincingly question whether the aim of achieving consistency in sentencing has been furthered.

Would removing some minimum sentences be viewed as diminishing the rights of victims?

The women’s rights lobby and victim groups support the extension of the minimum sentences legislation, especially insofar as it targets certain sexual offences and signals the severity of crimes involving sexual violence.

The submission of the Western Cape Consortium on Violence Against Women analysed an array of recent cases in support of their contention that the abolition (or non-extension) of the mandatory sentencing legislation would prejudice women. Among other arguments, the cases they cite reveal that, in their opinion, the following factors are consistently and erroneously used to justify lesser sentences, namely:

• the previous sexual history of the complainant;86

• an accused’s cultural beliefs about sexual assault;87

• an accused’s use of intoxicating substances prior to the assault;88

• an accused’s lack of intention to cause harm to the complainant in committing the rape;89

• an accused’s lack of education, sophistication or a disadvantaged background;90

• a lack of ‘excessive force’ used to perpetrate the rape;91

• a lack or apparent lack of physical harm to the complainant;92

• a lack or apparent lack of psychological harm to the complainant;93 or

• any relationship between the accused and the complainant prior to the offence being committed (including a consensual sexual relationship).94

The Consortium noted that the above factors were precisely those that had earlier been argued to be factors which the legislation should spell out as those that courts may not take into account in determining substantial and compelling circumstances. With this in mind the Consortium put forward the following suggestion as one optional reform proposal:

In light of the problematic jurisprudence on the meaning of ‘substantial and compelling circumstances,’ it is submitted that Parliament must enact mandatory interpretative guidelines for the judiciary, setting out how it is to be interpreted in light of the Constitution and international obligations to protect the rights and dignity of women. Specifically, the legislature should set out circumstances or factors that may not in themselves be regarded as ‘substantial and compelling circumstances ... .95

What impact has the minimum sentencing laws had on court procedure and efficiency?

As mentioned above, section 52 of the legislation allows for a split procedure where an accused may be tried in a regional court, and referred for sentence to a High Court; a process found to be constitutionally acceptable.96 However, concerns remain despite the Constitutional Court’s finding that the split procedure does not invade an accused’s fair trial rights, as any factor relevant to sentencing could still be placed before the sentencing court, even though this court had not heard the original evidence. Concerns include that the split procedure causes serious institutional delays and duplication of work, as trial records have to be typed up and studied again by the sentencing officer, new defence counsel and different prosecutors. Additionally, judges sometimes have to rehear witnesses, call for new evidence at sentencing stage, or refer matters back. Judges are also forced to exercise appeal-type scrutiny over convictions of regional courts – a role which some find improper.

The Western Cape Consortium on Violence against Women presents a slightly different view on the efficiency issue, proceeding from the perspective of the victim:

A further aspect that deserves urgent investigation is the extent to which original convictions are overturned by the High Court following referral for sentencing in terms of Act 105 of 1997 after trial in the regional court. In many cases, such action by the High Court results in the complainant having to testify again, thus doubling the secondary traumatisation that she may have experienced the first time round. At present, there is insufficient data to ascertain whether there may be trends of negligence (whether on the part of the prosecution or presiding officers) during the original trial, resulting in an unsupportable conviction, or whether High Courts are perhaps (with respect) being over-sensitive in assessing such convictions.97

Commentators have also drawn attention to the poor drafting of the Act, and the difficulties this has caused the courts as they have had to start delving into the details of the offence types set out in Schedule 2.98 Problems were also occasioned concerning the applicability of the legislation to district courts, as the law makes mention only of regional courts and high courts. Prof. Terblanche argues that this may have been justifiable in the light of the fact that the legislation was introduced as an emergency measure, but he suggests that it is now time for it to go:

One should not be fooled into believing that the Act is anything but an expensive tool. Just consider the many thousands of judicial officer hours that have been consumed in trying to make sense of its provisions, or trying to get around those provisions that turned out to be patently unfair … These costs might have been worthwhile if the Act had actually achieved its purpose.99

In the context of the cost implications of minimum sentencing laws, a further question has to be asked as to whether South Africa can afford a prison population of the size that it is now; and one that is bound to escalate, with prisoners serving ever longer terms of imprisonment. Not only are current overcrowding levels alarming, but the circumstances under which prisoners are accommodated may well be unconstitutional.100 We cannot build our way out of the problem, and the nexus between sentencing regimes and conditions of imprisonment need to be re-established urgently.101

How are the new parole provisions relevant to debates around mandatory sentencing?

The Correctional Services Act 111 of 1998 was promulgated during July and October 2004, and is now fully in force. New parole provisions designed to meet public criticism that prisoners are released too early and by an inappropriate process,102 are features of the new Act. In terms of the new legislation, all prisoners must serve at least half their sentences, and only after this may they be considered for conditional release on parole. Where a life sentence has been imposed, however, consideration of release may occur only after the expiry of 25 years; by comparison, the trend until now has been that persons serving life sentences may be considered for parole after having served 20 years of their sentence. It seems evident that the harsher parole provisions will inevitably result in prisoners spending longer in prisons, which may exacerbate overcrowding over the long term. This effect may be compounded when regard is had to the new provisions concerning all prisoners sentenced in accordance with Act 105 of 1997, as minimum non-parole periods have been set at two-thirds (and even four-fifths in some instances) of the sentence.

However, in the absence of concrete data indicating how many prisoners would be affected by these provisions, the actual effect of the new parole provisions upon the future inmate population is still uncertain.

Towards comprehensive sentencing reform

As pointed out, in 1998 a new project committee of the South African Law Commission was appointed by the Minister of Justice and headed by Prof. Dirk van Zyl Smit. The committee was tasked with continuing the research on sentencing reform, and also with giving consideration to the position of victims in the criminal justice system. As mentioned earlier, the work of this committee (hereafter referred to as the new committee) culminated in the development of a report, entitled Sentencing: A new sentencing framework, together with a proposed Sentencing Framework Bill released in 2000.103

In their Report, this new committee cites a number of criticisms levelled at the mandatory minimum sentences introduced by the 1997 Criminal Law Amendment Act. First, it reiterated the criticism by judicial officers that the provisions limited their discretion. Second, it was noted that the 1997 Act deals only with a limited number of crimes, while other serious crimes (such as kidnapping) are not included, thus disturbing the proportionality between various types of crimes. Lastly, it was reiterated that if the intention was to move towards consistency in sentencing, this had not been achieved given that judges had had difficulty in applying the ‘substantial and compelling circumstances’ test.

Taking cognisance of these criticisms, the new committee saw its task to be the development of a model for sentencing reform suitable for the local context. It was the view that an ideal system should promote consistency in sentencing, deal appropriately with concerns that particular offences are not being regarded with a suitable 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.104

In order to develop sound recommendations, empirical studies were undertaken by the new committee between June 1999 and January 2000, first on sentencing patterns before and after the introduction of the Criminal Law Amendment Act, and then on attitudes of the key role players towards the Act.105 Key empirical findings from this research were that there were significant disparities in sentencing for serious offences, particularly on regional lines, and that these persisted even after the coming into effect of the 1997 Criminal Law Amendment Act.106

Although this research was conducted shortly after the implementation of the legislation and may have been premature, indications were then that as far as the goal of achieving uniformity and consistency in sentencing was concerned, the mandatory minimum sentencing legislation was not having the desired effect.

The qualitative research that was conducted regarding the views of criminal justice role players on sentencing, indicated a wide range of opinions on sentencing practice. This included, among others, strong criticism of perceived inconsistencies in sentencing, and strong opposition (particularly among the judges) to the idea of mandatory minimum sentences.107 The committee also analysed the submissions to Parliament that were made at the time that the minimum sentencing legislation was being considered. According to the Report there was considerable divergence of views regarding the desirability of mandatory minimum sentences and, indeed, any attempt to limit sentencing discretion.108

The Sentencing report acknowledged the work done by the Van den Heever Committee in presenting the international sentencing models and does not revisit these in any detail. One significant development during the time that elapsed between the Issue Paper and the conclusion of the Report was the mechanism created for the development of sentencing guidelines in England and Wales. The Report summarises the development of what it terms this ‘hybrid model’ as follows:

Section 81 of the omnibus Crime and Disorder Act 1998 provides for the creation of an expert Sentencing Advisory Panel to advise the Court of Criminal Appeal on the new functions that it has in terms of Section 80. The essence of these functions is that whenever the Court of Appeal deals with an appeal against a sentence or when asked to do so by the Panel, it must consider formulating guidelines. Where such guidelines already exist the Court must consider whether it should review them.109

Section 80(3) provided further that where the Court decides to frame or revise such guidelines, a list of specific factors must be taken into consideration, including: the need to promote consistency in sentencing; actual sentences opposed by courts for offences of the relevant category; the cost of different sentences and their relative effectiveness in preventing offences; the need to promote public confidence in the criminal justice system; and the views communicated to the Court by the Sentencing Advisory Panel.110

At the time that the Sentencing report was completed, this envisaged Sentencing Advisory Panel was in its infancy; distinguished expert members were appointed only in mid-1999. Within a fairly short period of time it was noted that the Court of Appeal had adopted the Panel’s advice in a number of cases.111 However, the Court of Appeal had to wait for suitable cases to appear before it could issue guidelines, and then the guideline was confined to a particular offence or series of offences.112

Sentencing issues continued to dominate the public agenda in England and Wales, leading to further proposals. The so-called Halliday Report released by the Home Office proposed a Sentencing Guidelines Council chaired by the Lord Chief Justice, with members from the judiciary and a range of others with experience in the criminal justice system. The Sentencing Advisory Panel would remain in place, and the relationship between the Panel and the Council would in effect be the same as that between the Panel and the Court of Appeal. These proposals were formalised in sections 167 to 173 of the Criminal Justice Act 2003. The Sentencing Council was established in 2004 and has already issued guidelines on the discount on sentences to be offered for early guilty pleas, assessing the seriousness of crimes, and the implementation of new sentences.113

The Criminal Justice Act 2003 also introduced new, more onerous provisions on sentencing, which came into force in April 2005. For instance, a new sentence of imprisonment for public protection, with the characteristics of a sentence of imprisonment for life, has been created and extended to a great many offences (suggested by the Lord Chief Justice to be an excessive range).114 The scale of change is regarded as being so great (and, in addition, allegedly controversial among the judiciary) that it was decided that every judge engaged in sentencing should receive training at a residential course. Developments in England and Wales have therefore proceeded considerably since the publication of the South African Law Commission’s Report; consequently, the Report’s proposals, insofar as they were influenced by the situation prevailing then in England and Wales, should possibly be reviewed.

Prior to finalising its recommendations in the Report, the committee in April 2000 published a discussion paper on a new sentencing framework, and debated this proposal at workshops around the country. Most respondents agreed with the basic approach of the discussion paper that sentencing should be the outcome of a legislatively structured partnership between the various branches of government, and that normative sentencing guidelines should be established by a combination of a Sentencing Council and the Supreme Court of Appeal.115 However, some judges had objected to the idea of sentencing guidelines to be developed by a Sentencing Council, even if these allowed for a degree of flexibility in their application.116

Nevertheless, and in the absence of concrete proposals on sentencing reform from those quarters, the Report by and large reiterated the approach taken in the Discussion Paper, save in one important respect. The Discussion Paper had envisaged a combination of sentencing guidelines developed by a Sentencing Council, together with guideline judgments to be given by the Supreme Court of Appeal. The Report did not continue to assert a role for the Supreme Court of Appeal, citing as a reason the potential impracticality of a two tier model, and the lack of comprehensiveness that might result.

The Report therefore proposed that, as a start, sentencing principles should be clearly articulated in legislation. This heralded a decisive break with the common law approach, characterised by unstructured discretion. Supplementing this would be sentencing guidelines, developed by an independent Sentencing Council for a category or sub-category of offences. This Council would also have the mandate to collect and publish comprehensive sentencing data on an annual basis. Reports would have to: include commentary on the efficacy and cost effectiveness of the various sentencing options provided by legislation; determine the value of fine units; and make policy recommendations on the further development of community-based sanctions.117 The Council would be constituted in such a way as to allow the judiciary to have major input on the shaping of the guidelines themselves, but would involve role players from other branches of government (e.g. from the NDPP’s office, from the Department of Correctional Services, and from the magistracy). The role and determination of sentencing guidelines was detailed, as well as their application upon conviction. Provision was made for a 30% deviation upwards or downwards in the severity of a sentencing option, and the criteria for reasonable departures from the guideline was outlined.

Although this investigation was completed some five years ago, no evident movement to introduce the draft Sentencing Framework Bill contained in the Report (or something along similar lines) has occurred, despite the obvious and continued public interest in the sentencing process. It has been suggested that the Sentencing Framework Bill never formally entered the public arena due to two possible factors: one, the identified disapproval from some judges regarding the possible limiting of their discretion; and two, the fact that there was no co-ordinated strategy or response from civil society.118 In relation to some other recent Law Commission investigations (such as the Child Justice Bill 49 of 2002, the Sexual Offences Bill 50 of 2003, and the Children’s Bill 70 of 2003) there has been, in addition to solid law reform proposals, a strong civil society lobby aimed at ensuring the passage of legislation in Parliament. There has thus far been no similar initiative in relation to sentencing reform.

However, as the recent debates about the extension of Act 105 of 1997 appear to show, the climate in this regard may have changed. In her April 2005 constituency newsletter, Deputy Minister of Correctional Services Cheryl Gillwald says:

[a] review of our sentencing framework is perhaps more necessary than ever before. It is the belief of the Department that sentences should reflect the seriousness of the crime perpetrated – that goes without saying. A worrisome feature, however, of our criminal justice system seems to indicate that sentencing trends in our country show a marked difference to progressive and comparable jurisdictions internationally.119

There are other signs that a more fertile climate for considering sentencing reform may be on the cards, and there is an emerging shift in emphasis in the entire criminal justice system. The period during which the minimum sentences laws were passed were characterised by a ‘get tough on crime’, zero tolerance approach, evidenced by the introduction of harsher bail laws, the establishment of the Scorpions, measures to deal with organised crime, and so forth. However, it has been suggested recently that a longer-term vision regarding social and developmental approaches may be taking precedence over immediate short-term responses to crime.120 The upcoming Criminal Justice review will re-look at the approaches originally recommended by the NCPS, including restorative justice, diversion, and social crime prevention. This may well provide more fertile soil for debate around innovative sentencing proposals.

Conclusion

The stated intention of the legislature with the introduction of the minimum sentencing provisions in 1997 was to reduce serious and violent crime, to achieve consistency in sentencing, and to address public perceptions that sentences were not sufficiently severe.

The impact of minimum sentencing is difficult to quantify. However, as can be seen from the information presented in this paper, there has been little or no significant impact with regard to any of the above goals as a result of the introduction of minimum sentencing legislation. Regarding its deterrent function, no substantive claims have been made that crime has been reduced. Likewise, it would seem that the South African criminal justice system is no closer to achieving consistency in sentencing than it was in 1997.

In terms of addressing fear of crime and public perceptions with regard to sentencing for serious offences, the 2003 Institute for Security Studies (ISS) victim survey indicates that public feelings of safety have actually declined since 1998, with significantly more people feeling unsafe in 2003 than they did in 1998.121 Furthermore, indications are that a minimum sentencing regime has the potential to exacerbate the existing problem of overcrowding in South African prisons. It must be pointed out that there is also no guarantee that if the minimum sentences contained in Act 105 of 1997 were abolished, the sentencing tariff would drop. Indeed, ever sensitive to the public mood, sentencing officers are by and large unlikely to shift sentence terms measurably downwards.

It is therefore suggested that a more comprehensive sentencing reform initiative should be a matter of priority. This could be in the form envisaged in the Law Commission’s Report or in some other form, such as via guideline judgments of the Supreme Court of Appeal, or one of the other options proposed in the Issue Paper and outlined at the beginning of this paper. Possibly, the hesitance evidenced in the Law Commission’s Report to involve the Supreme Court of Appeal in setting sentencing guidelines should be reconsidered. And, with nine years of prescribed sentencing history already behind us, judges may now be more receptive to proposals concerning structured sentencing than they were formerly.

Against the backdrop of international developments in this area that are proceeding apace, and with a burgeoning prison population destined to spend longer and longer behind bars, future generations of South Africans are not going to thank us for our inaction and failure to grasp the nettle. The recent extension of Act 105 of 1997 might appear to have bought some time, but experience shows, too, that as far as parliamentary procedures are concerned, two years is all too short. Without clear and agreed plans on the table very shortly, it could well be too late.

Notes

1. See for example Sunday Tribune, 30 January 2005.
   
   

2. Sunday Tribune, 30 January 2005, quoting Democratic Alliance Spokesperson on Justice, Sheila Camerer. The death sentence was ruled unconstitutional in 1995 in S v Makwanyane and another 1995 2 SACR 1(CC).
   
   

3. Then called the South African Law Commission.
   
   

4. A J Ashworth, Criminal justice and deserved sentences, Criminal Law Review, cited in South African Law Commission, (Project 82) Sentencing; Mandatory minimum sentences, Issue Paper 11, 1997, p 4.
   
   

5. R G Nairn, Sentencing S v Young 1997 1 SA 602 (A) 1977 SACC, 189-191.
   
   

6. Ibid, 191.
   
   

7. Quoted in South African Law Commission, Issue Paper 11, op cit, par 2.2.
   
   

8. M A Rabie & S A Strauss, Punishment: An introduction to principles, 4th edition, Lex Patria, 1985, p 6, cited in South African Law Commission, Issue Paper 11, op cit, p 6.
   
   

9. S v Khumalo 1984 3 SA 327 (A) at 330 D-E.
   
   

10. South African Law Commission, Issue Paper 11, op cit, par 2.25.
    
    

11. S v Peters 1987 3 SA 717(A).
    
    

12. South African Law Commission, Issue Paper 11, op cit, par 2.27.
    
    

13. S v Toms; S v Bruce 1990 2 SA 802 (A) at 817 C-D.
    
    

14. S v Zinn 1969 2 SA 537 (A) at 540G.
    
    

15. South African Law Commission, Issue Paper 11, op cit, par 2.45.
    
    

16. Sentences of incarceration in which an offender is given a fixed term that may be reduced by good time or earned time.
    
    

17. South African Law Commission, Issue Paper 11, op cit, par 3.11.
    
    

18. Ibid, par 3.33.
    
    

19. Ibid, chapter 4.
    
    

20. Ibid, par 4.2.
    
    

21. Ibid, par 4.5.
    
    

22. Closing date for comment on the Issue Paper was the end of September 1997, but the legislation was already being considered in the National Assembly, the portfolio committee work having been finalised in November 1997.
    
    

23. South African Law Commission Report, Project 82, Sentencing: A new sentencing framework, 2000.
    
    

24. Given the rather bland title of the envisaged legislation, few were aware of the fact that the main issue was in fact the introduction of a new sentencing regime.
    
    

25. A Skelton, Juvenile justice reform: Children’s rights versus crime control, in C J Davel (ed), Children’s rights in a transitional society, Protea Books, 1999.
    
    

26. See S S Terblanche, Mandatory and minimum sentences: Considering s51 of the Criminal Law Amendment Act 1997, in J Burchell & A Erasmus (eds), Criminal justice in a new society, Juta and Co, 2003, p 195 quoting phrases such as ‘appallingly bad manner in which sections have been drafted’,’ill conceived and badly drafted’, ‘peculiar wording’ and ‘unhappily drafted legislation’ from an array of court cases.
    
    

27. See reference to murder in Part I (a), (b)(i), (c)(i), (c)(ii) and (d) of Schedule 2 to the Act.
    
    

28. See reference to rape in Part I (a)(i)-(iv) of Schedule 2 to the Act.
    
    

29. See reference to rape in Part I (b)(i)-(iii).
    
    

30. Part reference to rape in Part I (c) of Schedule 2 to the Act.
    
    

31. See Part II of Schedule 2, Part III of Schedule 2 and Part IV of Schedule 2.
    
    

32. Similar provisions apply to offences listed in Part III of Schedule 2 (10, 15 and 20 years) and Part IV of Schedule 2 (5, 7 and 10 years respectively for first, second and third offenders).
    
    

33. See, for example, D van Zyl Smit, Mandatory minimum sentences and departures from them in substantial and compelling circumstances, South African Journal on Human Rights 15, 1999, p 270.
    
    

34. Ibid, p 272.
    
    

35. Ibid.
    
    

36. R v Smith 34 CCC (3d) 97 (1987) and R v Vries 1996 (2) SACR 638 (Nm).
    
    

37. S v Mofokeng 1999 (1) SACR 502 (T), p 523c.
    
    

38. S v Malgas 2001(1) SACR 469 (SCA).
    
    

39. Ibid, par 18.
    
    

40. Ibid, par 19.
    
    

41. Ibid, par 20.
    
    

42. Ibid, par 25.
    
    

43. S v Dodo 2001 (1) South African Criminal Law Reports 593.
    
    

44. Section 35(3)(c) of the Constitution guarantees to every accused person ‘a public trial before an ordinary court’.
    
    

45. Section 12(1)(e) of the Constitution.
    
    

46. See Skelton, op cit. See too J Sloth-Nielsen, The influence of international law in South Africa’s juvenile justice reform process, unpublished LLD dissertation, University of the Western Cape, 2001.
    
    

47. Brandt v S Case 513/03, Supreme Court of Appeal.
    
    

48. This was the approach adopted by a preceding case, namely Direkteur van Openbare Vervolgings v Makwetsja 2004 (2) SACR 1 (T)), and by the majority of the judges on appeal in the lower court in Brandt.
    
    

49. Par 24(d) of the judgment.
    
    

50. S v Dodo, op cit, par 24.
    
    

51. Ibid, par 25-26.
    
    

52. Ibid, par 37. The Constitutional Court mentioned, too, that a punishment may violate the prohibition against the imposition of punishment which is cruel, inhuman and degrading for other reasons, for example, due to its nature or the conditions under which it is served.
    
    

53. S v Dodo, op cit, par 38.
    
    

54. See too H Fagan, Curb the vengeance – laws on minimum sentencing and parole spell worsening prison conditions, SA Crime Quarterly 10, December 2004.
    
    

55. At a briefing to the Supreme Court of Appeal, September 2004.
    
    

56. Sunday Tribune, 30 January 2005. It has also been asserted that an increase in the sentencing jurisdiction of the lower courts has played a contributory role in escalating prison terms, and consequently overcrowding. The jurisdiction for district courts was increased to three years, and that of regional courts from ten to 15 years. However, proper case-by-case studies would be required to ascertain whether the tariff has indeed escalated as a consequence of increased sentencing jurisdiction.
    
    

57. J Steinberg, Prison overcrowding and the constitutional right to adequate accommodation in South Africa, Occasional Paper, Centre for the Study of Violence and Reconciliation, 2005.
    
    

58. Ibid.
    
    

59. Ibid, p 10 quoting D Gilliard & A Beck, Prisons and jail inmates, Bureau of Justice Statistics, Washington DC, 1996; J Woolredge, A state level analysis of sentencing policy and inmate overcrowding, Crime and delinquency 42(2), 1996, p 456; T Gabore & N Crutcher, Mandatory minimum penalties: Their effect on crime, sentencing disparities and criminal justice expenditure, Canadian Department of Justice, Research and Statistics Division, Ottawa, 1997.
    
    

60. W Hartley, Jails may soon turn away prisoners, Business Day, 14 April 2005.
    
    

61. See South African Law Commission, Conviction rates and other outcomes of crimes reported in eight South African police areas, Research Paper 18, 2000. This study concluded, somewhat bleakly, that crime does indeed pay, with only six convictions after more than two years for every 100 violent crimes (murder, rape and aggravated robbery). Less than one in 20 perpetrators of reported adult rape (5%) or aggravated robbery (3%) is convicted. Moreover, South Africa’s figures did not appear to fare well compared to other countries for which comparable data was available. In those jurisdictions, approximately 50% of reported murders result in convictions, compared to 11% in South Africa. The conviction rate for reported rape is 19% in the US and 10% in England and Wales, compared to 7% in South Africa (Research Paper 18, p 26).
    
    

62. South African Law Commission, Project 82, Empirical study of the sentencing practices in South Africa, Research Paper 17, 2000.
    
    

63. M Tonry, Sentencing, judicial discretion and training, 1992. Quoted in S S Terblanche, Mandatory and minimum sentences: Considering s 51 of the Criminal Law Amendment Act 1997, Acta Juridica 194, 2003, p 137.
    
    

64. The National Assembly and the National Council of Provinces have by resolution, on 12 and 13 April 2005, voted to support a further extension of the Act until 30 April 2007. The Proclamation by the State President to give effect to this resolution has been signed. It is perhaps worthy of note that the parliamentary extension was supported by all political parties.
    
    

65. A three year moratorium (imposed in July 2000) on the release of crime statistics was lifted on 22 September 2003.
    
    

66. W Scharf & J Berg, Crime statistics in South Africa 1994–2003, 17(1), South African Journal on Criminal Justice 57, 2004.
    
    

67. Ibid, p 61.
    
    

68. Ibid, p 67.
    
    

69. Ibid, pp 72-73.
    
    

70. A Altbeker, Puzzling statistics: Is South Africa really the world’s crime capital?, SA Crime Quarterly 11, 2005.
    
    

71. Ibid.
    
    

72. Part II (b) of Schedule 2.
    
    

73. See OSF-SA Minimum sentences roundtable report, March 2005, <www.osf.org.za>.
    
    

74. The point was also made by a member of the Department of Correctional Service at the Open Society Foundation Roundtable workshop on minimum sentences legislation that such prisoners have nothing to lose and are more willing to use violence and weapons in attempting to escape.
    
    

75. OSF-SA Minimum sentences roundtable report, op cit, p 7.
    
    

76. Referring to A Blumstein, Prisons, in J Q Wilson & J Peterseia (eds), Crime, ICS Press, California, 1995, p 408.
    
    

77. OSF-SA Minimum sentences roundtable report, op cit.
    
    

78. S S Terblanche, Sentencing guidelines for South Africa: Lessons from elsewhere, South African Law Journal 858, 2002, p 859.
    
    

79. Ibid, p 881. See, too, OSF-SA Minimum sentences roundtable report, op cit, p 12.
    
    

80. See Sunday Tribune, 30 January 2005 for allegations in this regard. It has been rumoured, further, that all of the judge presidents of the various provincial divisions of the High Courts provided submissions to the Department of Justice and Constitutional Development prior to the extension of the legislation, expressing their (and their benches’) opposition to the continuation of the life of Act 105 of 1997.
    
    

81. The authors are indebted to Helene Combrinck, Senior Researcher, Gender Project, Community Law Centre, University of the Western Cape, for making this submission available to us.
    
    

82. S v Mahomotsa 2002 (2) SACR 435 (SCA).
    
    

83. S v Abrahams 2002 (1) SACR 116 (SCA).
    
    

84. Ibid, para 18, emphasis added in original submission.
    
    

85. S v G 2004 (2) SACR 296 (W).
    
    

86. S v Mahamotsa, op cit.
    
    

87. S v Mvamvu (case 350/2003 (SCA)).
    
    

88. S v Njikelana 2003 (2) SACR 166 (C).
    
    

89. S v Mvamvu, op cit.
    
    

90. S v Njikelana, op cit.
    
    

91. S v G, op cit and S v Shongwe 1999 JDR 0473 (O).
    
    

92. S v G, op cit, S v Njikelana, op cit and S v Shongwe, op cit.
    
    

93. S v Mvamvu, op cit and S v Shongwe, op cit.
    
    

94. S v Mvamvu, op cit and S v Abrahams, op cit.
    
    

95. Submission of the Western Cape Consortium on Violence Against Women, (copy on file with authors), 2005, p 7.
    
    

96. S v Dzukada, S v Tilly, S v Tshilo 2000 (2( SACR 443 (CC).
    
    

97. Submission of the Western Cape Consortium on Violence against Women, op cit, p 11.
    
    

98. Terblanche, 2003, op cit, p 194.
    
    

99. Ibid, p 220.
    
    

100. Steinberg, op cit.
     
     

101. Steinberg (ibid, p 13) notes that this link appeared to have fallen off the judicial map, and refers in this regard to a study in 2002 in which 42 magistrates and High Court judges were asked whether the capacity of the correctional system to carry out sentences should be considered when sentence is imposed: 80% of the respondents said never or almost never, and only 10% said always or almost always (M Schonteich, D Mistry & J Struwig, Qualitative research report on sentencing: An empirical, qualitative study on the sentencing practices of the South African criminal courts, with a particular emphasis on the Criminal Law Amendment Act 105 of 1997, South African Law Commission, Discussion Paper, May 2000).
     
     

102. South African Law Commission Report, op cit, p 6.
     
     

103. This Report followed on a preceding Discussion Paper (No 91).
     
     

104. South African Law Commission Report, op cit.
     
     

105. South African Law Commission, An empirical quantitative and qualitative study of the sentencing practices of the South African Criminal courts, with particular emphasis on the Criminal Law Amendment Act, 105 of 1997, Research Paper 17, June 2000.
     
     

106. South African Law Commission Report, op cit, par 1.19.
     
     

107. Ibid, p 10.
     
     

108. Ibid, par 1.39.
     
     

109. Ibid, p 25.
     
     

110. Section 80(3) of the Crime and Disorder Act 1998.
     
     

111. Terblanche, 2002, p 858 at 871.
     
     

112. Lord Woolf, Lord Chief Justice, Making sense of sentencing, paper presented at the Sir Leon Radzinowicz Lecture, Cambridge Institute of criminology, 12 May 2005, <www.dcs.gov.uk/ judicial/speeches/lcj120505.htm> (17 May 2005).
     
     

113. Sentencing Guidelines Council, 16 December 2004 (see, further <www.setencing-guidelines.gov.uk>).
     
     

114. Lord Woolf, op cit.
     
     

115. South African Law Commission Report, op cit, par 1.44.
     
     

116. Ibid, par 1.47.
     
     

117. Ibid, par 2.9.
     
     

118. This was put forward at the Minimum sentences roundtable, Open Society Foundation, 30 January 2005.
     
     

119. Constituency Newsletter 52, April 2005, p 2.
     
     

120. A du Plessis & A Louw, Crime and crime prevention in South Africa: 10 years after, Canadian Journal of Criminology, April 2005.
     
     

121. P Burton, A du Plessis, T Leggett, A Louw, D Mistry & H van Vuuren, National victims of crime survey: South Africa 2003, ISS Monograph Series 101, Pretoria, July 2003.

About this paper

The Criminal Law Amendment Act 105 of 1997 introduced prescribed sentences for specified serious offences into the South African legislative framework. The period preceding Parliament’s third extension of this legislation in April 2005 saw intense lobbying, advocacy, and public debate about the efficacy and desirability of minimum sentences for serious crimes. This paper examines some of the arguments that have been raised for and against the present sentencing regime, and highlights the need for comprehensive sentencing reform in South Africa. The paper examines some key questions: Are mandatory minimum sentences constitutional? Have they deterred or prevented crime? Do they afford better protection to victims? What is the relationship between minimum sentences and prison overcrowding? Finally, the paper questions whether South Africa needs a more comprehensive sentencing reform strategy.

About the authors

Julia Sloth-Nielsen is a professor in the Law Faculty, University of the Western Cape, and research associate at the Community Law Centre, University of the Western Cape. She is a co-founder of the Civil Society Prison Reform Initiative (CSPRI); an organisation which promotes good governance in prisons, improved delivery of human rights to prisoners, and increased civil society involvement in and access to information about prisons.

Louise Ehlers is a senior project officer with the Criminal Justice Initiative of the Open Society Foundation for South Africa, and convenor of the Roundtable on Minimum Mandatory Sentences held on 31 January 2005.

Funder

This paper is funded by the Hanns Seidel Foundation, the Royal Danish Embassy, and the Ford Foundation.